DEV Community: DevToolsPicks

How to Use Claude Code as a Solo Developer in 2026

DevToolsPicks — Sat, 18 Apr 2026 05:54:14 +0000

You install Claude Code. You type claude in your terminal. Now what?

Most solo devs use maybe 10% of what Claude Code can actually do. They get stuck in a chat loop, paste code back and forth, and wonder why it feels like a fancy autocomplete. The problem isn't the tool. It's that most guides teach features one at a time instead of showing how they fit together in a real shipping workflow.

This post covers the opposite approach. Not a feature catalog. Not a full tutorial. Just the workflow that actually works for solo developers shipping production software in 2026, no matter what stack you're on.

Start with CLAUDE.md. Nothing else matters as much.

The single highest-ROI setup step is a well-written CLAUDE.md at the root of every project.

Claude Code reads this file on every session. If it's good, Claude behaves like it already works on your codebase. If it's missing or vague, Claude will guess. And when it guesses wrong, it will hand you patterns from a different framework, suggest raw SQL where you have an ORM, or miss that you're using a UI library where it assumed plain components.

Here's the structure of a good CLAUDE.md:

# Tech stack
TypeScript + Next.js 15 + Postgres. Deployed on Vercel.
Authentication via Clerk. Payments via Stripe.

## Code style
- TypeScript strict mode. No `any`. Prefer `unknown` with type guards.
- Run the linter before every commit. Errors, not warnings.
- Functional components only. No class components.

## Don't
- Don't drop tables or run destructive migrations without asking first.
- Don't add comments that explain what code does. Only why.
- Don't install a new dependency without flagging it.

## Project conventions
- API routes live in app/api/. Server actions in lib/actions/.
- All validation uses zod schemas, never manual checks.
- Components get extracted after 3+ uses. Inline before that.

Notice what's not in there: architecture diagrams, philosophy, full style guides. CLAUDE.md is for the things Claude will otherwise get wrong, not for documenting your project end-to-end. Two clear lines that stop a destructive migration beat ten paragraphs about hexagonal architecture.

Adapt the stack and rules to match your own project. The structure is what matters: stack summary, code style rules, things to never do, project conventions. That's it.

For advanced memory management (.claude/rules/ for path-specific overrides, CLAUDE.md precedence across repos), the Claude Skills vs MCP Connectors vs Plugins breakdown explains how this fits into the broader system. CLAUDE.md is where to start though.

The three modes that matter

Claude Code has six permission modes. In practice, three of them get real use.

Plan mode (/plan or shift-tab). For anything that touches more than one file. Plan mode stops Claude from rushing to write code. Instead it produces a step-by-step plan you can push back on before a single line gets written. It saves a ridiculous amount of time. Catching a bad plan before Claude refactors something you didn't want touched takes thirty seconds. Cleaning up after a bad refactor takes an hour.

Default mode (acceptEdits with review). For focused single-file work. Fix a bug, add a method, refactor a function. Default mode shows each change before applying it. Fast feedback loop.

Auto mode. New with Opus 4.7 for Max plan users. Claude makes its own decisions about when to search files, when to run tests, when to verify changes. Useful for background batch work like running a linter fix pass across a repo or generating boilerplate for several components at once. Not ideal for architectural work where you still want to review decisions, not just outcomes. The Opus 4.7 launch review covers what else changed with this release.

One mode to avoid: bypassPermissions. It sounds fast. It also means Claude can run destructive commands without asking. A solo dev who just lost a day to a bad rm -rf is not a solo dev shipping anything that week.

Slash commands worth building yourself

The built-in slash commands are fine. The ones that actually move the needle are the ones you build for your specific stack and the specific mistakes that waste your time.

Three patterns worth copying:

/lint-fix. Runs the linter on the current file, then asks Claude to review what the linter couldn't catch: unused imports that slipped through, return types that should be stricter, error handling that's too broad, type assertions masking real bugs. Kills the loop of "run lint, review diff, ask Claude, repeat."

---
name: lint-fix
description: Run the linter then catch what the linter misses
---

1. Run the linter on the current file
2. Review the diff for:
   - Type assertions that could be replaced with proper types
   - Return types that are implicit but should be explicit
   - Unused imports the linter did not catch
   - Error handling that is too broad (catch-all patterns)
3. Show suggested changes separately from the lint output
4. Do not auto-apply type changes. Show the diff and wait.

/migration-review. Scans a just-generated database migration for gotchas. Foreign key constraints pointing the wrong way. Missing indexes on foreign keys. Columns that should be nullable. Enum constraints that don't match application code. AI tools generate migrations well most of the time. The one time they miss an index on a foreign key, the cost shows up at scale.

/test-one. Bootstraps a single test for a specific function or method. Just the happy path plus one edge case. Asking Claude to "write tests" produces 400-line test files nobody maintains. Asking for exactly two cases produces tests that ship.

The common thread: slash commands that match your stack and your specific pain points. Generic helpers already exist. The ones you build yourself are the ones that get used every day.

Which model to pick, and when

Claude Code lets you pick between Opus 4.7, Sonnet 4.6, and Haiku 4.5. The honest split looks like this:

Opus 4.7 for architecture, multi-file refactors, debugging something subtle, and anything where being wrong costs an hour. It's $5/$25 per million tokens on the API, but on Max plans it's included in the flat rate. On Opus 4.7 specifically, the new /ultrareview command is worth running before shipping anything that touches auth or payments.

Sonnet 4.6 for about 80% of daily coding. Routine feature work, adding a column, writing a controller method, updating a component. Cheaper, faster, and good enough that the quality difference only shows up maybe one time in five. The Cursor vs GitHub Copilot vs Claude Code comparison goes deeper on this if the choice is between tools rather than models.

Haiku 4.5 for documentation, commit messages, quick lookups, and running the same prompt against ten files at once. Haiku is fast enough that it stops feeling like an LLM. Bad choice for reasoning work though. It will confidently make things up.

The common mistake is defaulting to Opus for everything. It works, but Pro plan rate limits blow through in three days. Better default: start on Sonnet. Escalate to Opus when Sonnet gets something wrong twice.

Pro vs Max vs API: the honest call

This is the most common question about Claude Code. Here's the decision framework by actual usage pattern.

Claude Pro at $20/month. If Claude Code is a supplement to your workflow, not the main thing. Maybe one to two hours of active use a day. Rate limits will bite occasionally, especially on Opus. Pro is the right tier for the first two or three months of using Claude Code seriously.

Claude Max 5x at $100/month. If Claude Code is your primary development tool. Active sprints typically hit 50 to 80 million tokens a month, which would cost $800 to $1,500 at API rates. Max 5x pays for itself inside one month of real use. No rate limit anxiety is worth the upgrade on its own.

Claude Max 20x at $200/month. For parallel sessions across multiple repos, heavy use of the Claude Code desktop app's parallel sessions feature (covered in the Claude Code desktop redesign review), or long-horizon autonomous agents. Most solo devs don't need this tier. Anyone who does already knows.

API direct billing. Only relevant for building something on top of Claude Code, not for using it as a developer. At API rates, a typical heavy day of coding costs more than a full month of Max 5x. The math doesn't work for personal use.

If the call is also between Claude Max and ChatGPT Pro $100, the ChatGPT Pro $100 vs Claude Max vs Cursor breakdown covers the full three-way decision.

Four mistakes that cost real time

Letting sessions grow too long. Claude's context window is big but not infinite. A three-hour session where new tasks keep piling on starts producing weird mistakes around hour two. The fix is simple: end the session, summarise the state in CLAUDE.md or a task file, start fresh. Pushing through context bloat instead of resetting costs hours and degrades output quality in ways that are hard to spot in the moment.

Skipping plan mode on "quick" changes. Anything that touches three or more files deserves plan mode, even if it looks simple. A "quick" refactor without plan mode can easily touch fourteen files. Cleaning up afterwards takes longer than writing the whole thing by hand would have. Plan mode isn't for slow jobs. It's for jobs where knowing what's about to happen matters.

Trusting generated tests without reading them. Claude writes tests that pass. That's not the same thing as tests that catch bugs. Every generated test needs a read before acceptance. Often the assertion tests the wrong thing, or the setup is missing a case, or the test passes even when the code is broken because a mock is wrong. This matters more than most people admit.

Under-using hooks. A pre-commit hook that runs the affected tests before any commit reaches git is maybe the single highest-ROI thing to add to a Claude Code workflow. Claude generates code fast. A hook that catches bad code before it reaches git history is the safety net that lets you trust the speed.

When NOT to reach for Claude Code

There are scenarios where Claude Code is the wrong answer.

One-line changes. Fixing a typo, adjusting a constant, renaming a variable. Opening Claude, waiting for a response, reviewing a diff, applying it. That whole loop takes longer than typing the change.

Architectural decisions that haven't been thought through. Asking Claude Code "should I use service classes or action classes for this?" produces a confident answer. It might be a good one. There's no way to tell without already having a view. Form the view first. Use Claude to pressure-test it, not to generate it.

Debugging where you're the domain expert. If the bug is in code Claude wrote, sure, let it debug. If the bug is in your understanding of your own product's business logic, Claude will happily generate plausible-sounding fixes that make the symptom go away while the actual bug stays. Domain debugging stays with you.

Anything involving payments or auth at the core level. Claude Code works well for auth-adjacent work: email verification flows, password reset pages, session lifetime config. Poor choice for designing the auth system itself. The consequences of a subtle mistake are too high. Same goes for billing logic.

Claude Code is a force multiplier, not a substitute for judgment. Most of the people who get burned by it are skipping the judgment part.

FAQ

Do you need the Max plan to use Opus 4.7?

No. Claude Pro at $20/month includes Opus 4.7 access, just with lower rate limits. Max plans give significantly more Opus usage per 5-hour window, which matters when Claude Code is a primary dev tool.

Can Claude Code and Cursor be used together?

Yes, they coexist fine. Cursor handles IDE-integrated inline suggestions well. Claude Code handles multi-file agentic work in a terminal-first flow. Different strengths, both valid in the same workflow.

What's the difference between the Claude Code CLI and the desktop app?

Same backend, different interface. The CLI is keyboard-driven and fits a terminal-first workflow. The desktop app adds a visual interface with parallel sessions, drag-and-drop panes, and an integrated terminal. Full breakdown in the Claude Code desktop redesign review.

How do you stop Claude Code from running destructive commands?

Avoid bypassPermissions mode. Stick with default or acceptEdits, which require approval for anything that modifies the filesystem. Add a pre-commit git hook as a second safety layer. For auto mode, configure .claude/settings.json with explicitly forbidden commands.

Is Claude Code worth it for developers who also use ChatGPT or Gemini?

For primary coding work, yes. Claude Code's terminal-first design and the quality of Opus 4.7 on multi-file refactors are tuned for software engineering in a way general-purpose chat interfaces aren't. Pair it with whatever else works for non-coding tasks.

Final take

Claude Code isn't something you learn once and move on from. It pays back more the more time you put into the setup. Three things separate the solo devs who get real value from it from the ones who don't: they write a CLAUDE.md for every project, they use plan mode for multi-file work, and they build a handful of slash commands that fit their stack. That's the 20% that delivers most of the value.

Everything else in the feature list (hooks, subagents, MCP servers, skills) is worth learning too. Just do them in that order. The foundation has to work before anything else matters.

Codex Just Became a Full AI Agent: Is It Worth Switching From Claude Code?

DevToolsPicks — Fri, 17 Apr 2026 08:53:05 +0000

On April 16, one hour after Anthropic launched Claude Opus 4.7, OpenAI shipped a Codex update that went way beyond model weights. Codex now sees your Mac screen, clicks around in native apps, runs 90+ plugins, generates images inline, and remembers your preferences across sessions. OpenAI is no longer pitching Codex as a terminal coding tool. It is pitching it as a full desktop AI agent that happens to do code really well.

The tweet that summed up the day had 1.6M views and read: "OpenAI just stole the spotlight one hour after Opus 4.7 dropped. Anthropic bets on raw model power, OpenAI wants your whole workflow." That is the story.

So if you are currently using Claude Code and watching your X feed light up with Codex demos, the question is: should you switch? Here is an honest breakdown of what actually changed, where each tool still wins, and what it means for indie hackers and solo devs.

Quick Verdict

Question	Answer
Biggest Codex change today	Computer Use on macOS, 90+ plugins, persistent memory
Best for terminal-native agentic coding	Claude Code (still)
Best for desktop automation across apps	Codex (new)
Cheapest entry tier	Both $20/mo (ChatGPT Plus vs Claude Pro)
Best $100 tier value for heavy use	Close call, see breakdown below
EU/UK availability	Claude Code: yes. Codex Computer Use: no at launch
Worth switching if you are happy with Claude Code	Probably not yet

What Actually Changed in Codex Today

OpenAI announced the update on April 16, 2026, and it is the biggest Codex release since the Mac app launched in February. Here is what shipped.

Computer Use on macOS. Codex can now operate Mac apps with its own cursor. It sees what is on your screen, clicks buttons, types into fields, and completes tasks in native apps that do not expose APIs. OpenAI emphasizes this runs in the background so multiple Codex agents can operate your Mac in parallel without interfering with your own work. The feature is built by the team OpenAI acquired from Sky Applications (the original Apple Shortcuts/Workflow team). It is not available in the EEA, UK, or Switzerland at launch.

In-app browser. A built-in browser powered by OpenAI's Atlas technology. You can open local files or public pages that do not require sign-in, comment directly on the rendered page, and ask Codex to act on your feedback. OpenAI plans to expand this into full browser control over time.

Image generation. The gpt-image-1.5 model is integrated directly into Codex. No more switching to ChatGPT to generate UI mockups, diagrams, or design assets.

Persistent memory. Codex remembers preferences, workflows, tech stacks, and corrections across sessions. So you stop re-explaining your stack every morning.

90+ new plugins. Jira (via Atlassian Rovo), CircleCI, GitLab Issues, Microsoft Suite, Neon by Databricks, Slack, Notion, Google Workspace, and MCP servers. OpenAI says the curated plugin approach is a deliberate response to security concerns in more open agent platforms.

Long-running automations. Codex can reuse existing conversation threads, schedule future work for itself, and wake automatically to continue tasks across days or weeks.

Other improvements. Multiple terminal tabs, SSH into remote devboxes (in alpha), rich file previews in a sidebar for PDFs, spreadsheets, slides, and docs, a summary pane for plans and artifacts, GitHub review comment addressing, and first-time support for Intel Macs.

Under the hood, GPT-5.4 is the recommended Codex model with an experimental 1 million token context window. OpenAI says 3 million developers now use Codex weekly.

Computer Use: The One Feature Claude Code Does Not Have

This is the single biggest functional difference after today.

Claude Code is terminal-native. It operates on your codebase through the CLI, reads and writes files, runs shell commands, and works with any editor. It does not click around in GUI apps.

Codex now does both. It operates your codebase through the CLI just like Claude Code, and it can also drive native Mac apps: Xcode simulators, design tools, browser-based internal dashboards, app testing flows, anything that does not expose an API.

For most backend and web development, this does not matter. You are in your terminal and your editor, you are not clicking around native apps. Claude Code remains equal or better there.

For some specific workflows, Computer Use is genuinely a step-change:

Native iOS and macOS development. Testing flows in Xcode simulators, capturing screenshots, iterating on Swift UI changes by actually clicking the running app
Non-API-exposed tools. Internal dashboards, legacy apps, or admin panels where you would otherwise be writing Selenium scripts
Design iteration. Pushing code changes, opening the rendered page in the in-app browser, commenting on what needs to change, letting the agent iterate

If you do any of these daily, Codex is now the clear winner on capability. If you do not, the feature is mostly marketing.

Codex vs Claude Code: Head to Head

Developer experience

Claude Code is a CLI tool that runs in any terminal. You stay in your editor of choice (VS Code, JetBrains, Zed, Neovim, whatever). It reads your codebase, makes multi-file edits, and runs shell commands. The IDE integrations for VS Code and JetBrains add side-panel conversations without replacing your editor.

Codex now has three surfaces: the CLI, the IDE extensions, and the Codex macOS app. The Mac app is where the new features live. It includes worktrees so multiple agents can work the same repo in parallel without conflicts. If you work in a single terminal or single editor window today, the Codex app is a different mental model from Claude Code. Not worse, just different.

Winner: Tie. Claude Code wins if you want editor-agnostic terminal simplicity. Codex wins if you want a dedicated agent workspace with built-in multi-agent orchestration.

Model quality

GPT-5.4 (Codex default) has an experimental 1M token context window. Claude Opus 4.7 launched yesterday with its own benchmark gains but a 200K standard context window on consumer plans (500K on Enterprise Claude).

For raw coding benchmarks, both are close enough that real-world preference often comes down to individual tasks. Anthropic has historically led on multi-file refactors and deep codebase understanding. OpenAI has historically led on speed and execution. Opus 4.7 and GPT-5.4 both claim improvements in those respective areas. We reviewed Opus 4.7 specifically here if you want the benchmark breakdown.

Winner: Roughly tied. Task-dependent. Run your own test on a real task in your codebase before switching.

Pricing

Both companies ended up at the same pricing structure, which is not a coincidence.

Tier	OpenAI (Codex included)	Anthropic (Claude Code included)
Entry paid	Plus: $20/mo	Pro: $20/mo ($17 annual)
Middle	Pro: $100/mo (5x Plus limits)	Max 5x: $100/mo (5x Pro limits)
Top	Pro: $200/mo (20x Plus limits)	Max 20x: $200/mo (20x Pro limits)

OpenAI launched the $100 Pro tier on April 9 explicitly to match Anthropic's Max 5x. Through May 31, 2026, the new $100 Codex plan gets a temporary 10x Plus boost instead of 5x.

The price-per-tier is identical. The usage you get inside each tier depends on how heavy your daily sessions are and which tool's rate limits hit you harder. Reports from heavy Claude Code users on Max 20x still hitting rate limits on long days are common. Codex users report similar friction on Plus. For indie hackers running 4-6 hours of agent work daily, neither $20 tier is actually enough and both $100 tiers feel tight.

Winner: Tie on paper. On real usage, both leak. Pick based on which model you prefer, not which is cheaper.

Ecosystem and integrations

Codex today shipped 90+ plugins, bringing it above 100 total when combined with existing ones. The curated ecosystem includes Atlassian Rovo for Jira, CircleCI, GitLab, Slack, Notion, Microsoft Suite, Google Workspace, Neon, and MCP servers.

Claude Code has MCP server support and a growing ecosystem through Anthropic's marketplace launched in March 2026 (with Snowflake, Harvey, Replit as launch partners). The MCP protocol is open and has more community-built servers than OpenAI's curated plugin list.

Winner: Codex wins on curated breadth for today's announcement. Claude Code wins on the open protocol (MCP) where developers can self-build integrations.

Regional availability

Claude Code works in the EU, UK, Switzerland, and everywhere else Anthropic operates.

Codex Computer Use is explicitly not available in the EEA, UK, or Switzerland at launch. Codex personalization and memory features are still rolling out for Enterprise, Edu, and users in the EU and UK.

If you are an indie hacker in Europe (or a Turin-based developer like me), Codex Computer Use is off the table entirely right now.

Winner: Claude Code, decisively, if you are in the EU or UK.

Who Should Actually Switch

Switch to Codex if:

You do native macOS or iOS development and the Computer Use + Xcode simulator workflow is a clear upgrade
You want a dedicated multi-agent workspace rather than a terminal-based approach
You already live inside the OpenAI ecosystem (ChatGPT Plus subscribers who use it heavily)
You need the 90+ plugins out of the box and do not want to self-configure MCP servers

Stay on Claude Code if:

You are in the EU, UK, or Switzerland (Computer Use is blocked, which kills the main differentiator)
Your workflow is web/backend development inside a terminal and you do not need desktop automation
You prefer the open MCP protocol over curated plugins
You are on a Max 20x plan and getting real value from the higher Claude Code rate limits
You value model quality on long multi-file refactors and believe Anthropic's models still edge ahead there

Test both if:

You are a solo developer or indie hacker with no strong existing preference
You have $20/mo to spare for a month of testing the other platform
You are about to start a new project and can evaluate both against the same real task

Honest Cons of the New Codex

Every Codex launch review glossed over the cons. Here are five worth naming.

Mac-only Computer Use. Windows users get the Codex app with plugins and everything else, but not Computer Use. Linux users get none of it. If your main machine is not a Mac, the headline feature of today's launch is not for you.

Regional lockouts. Computer Use is blocked in the EEA, UK, and Switzerland. Memory and personalization are also staggered for those regions. Half of Europe cannot use the announcement's biggest features today.

Platform lock-in risk. OpenAI's superapp strategy (per WSJ reporting) is to unify ChatGPT, Codex, and Atlas browser into one product. That means deeper OpenAI ecosystem lock-in over time. If you want tool-agnostic flexibility, terminal-native Claude Code has less of this problem.

The pricing rebalance. On April 9, OpenAI "rebalanced" the Plus plan's Codex allocation to "support more sessions throughout the week instead of longer sessions in a single day." Translation: if you were running heavy Codex sessions on $20 Plus, you may now hit limits faster and be nudged toward the $100 tier. Claude Code has done similar rebalancing at $20 historically. Neither company is above this move.

Curated plugin ecosystem. OpenAI frames the curated approach as a security win, which is reasonable. It also means if a plugin you need is not in the 90+ list, you wait for OpenAI to add it or you use MCP (which Codex does support). Claude Code's MCP-first approach is more open at the cost of some security guardrails.

Pricing Table

Plan	Monthly	Claude Code usage	Codex usage	Notes
ChatGPT Plus	$20	N/A	1x baseline	OpenAI side of $20 tier
Claude Pro	$20 ($17 annual)	Full Claude Code access	N/A	Anthropic side of $20 tier
ChatGPT Pro $100	$100	N/A	5x Plus (10x until May 31)	OpenAI mid-tier
Claude Max 5x	$100	5x Pro limits	N/A	Anthropic mid-tier
ChatGPT Pro $200	$200	N/A	20x Plus	OpenAI top tier
Claude Max 20x	$200	20x Pro limits, priority access	N/A	Anthropic top tier
API pay-per-token	varies	Sonnet 4.6: $3/$15 per 1M, Opus 4.7: $5/$25 per 1M	GPT-5.4: varies by token type	Best for unpredictable usage

Both ecosystems top out at $200/mo for consumer plans. Above that, you are in Business or Enterprise territory with custom pricing.

FAQ

Is Codex now better than Claude Code overall?

No. It is better at new things Claude Code does not do (Computer Use on Mac, curated plugin ecosystem, in-app browser, image gen). Claude Code remains competitive or better at terminal-native agentic coding, multi-file refactors, and EU/UK accessibility. The right answer depends on your workflow.

Can I use Codex Computer Use in the EU?

Not at launch. OpenAI explicitly blocks Computer Use in the European Economic Area, the UK, and Switzerland. The rest of Codex (the app, the CLI, most plugins) works in those regions, but the headline feature from today's launch is geofenced.

Does Codex have MCP support?

Yes. Codex supports MCP servers alongside its curated plugin ecosystem. The difference is curation: OpenAI vets the 90+ plugins in its directory, while MCP servers are community-built and self-configured.

Which is cheaper for heavy users?

At the $100 and $200 tiers, they are priced identically. Real usage varies by tool. Claude Code users on Max 20x report hitting rate limits on long days. Codex users on Plus report the same. Neither is meaningfully cheaper in real-world heavy use.

Should I cancel my Claude Code subscription and switch to Codex?

Probably not today. The EU/UK availability issue, the Mac-only Computer Use limitation, and the fact that Claude Code is still strong at its core use case (terminal-native agentic coding) all argue for staying put unless one of the new Codex features specifically changes your workflow. We compared all three major AI coding subscriptions here if you want the full context.

What about Cursor 3?

Cursor 3 shipped parallel agents and a redesigned composer window on April 2, 2026. It competes with both Codex and Claude Code on the IDE-integrated side of the market. If you want a polished IDE-first experience rather than a CLI or separate agent app, Cursor 3 is still a strong option. It does not do Computer Use like Codex now does.

What about Claude Code's new desktop app?

Anthropic redesigned the Claude Code desktop app with parallel sessions and an integrated terminal earlier this month. We reviewed that redesign here. It narrows the UX gap with the Codex app but does not match Codex Computer Use.

Final Recommendation

For most indie hackers and solo developers, the honest answer is do not switch yet.

If you are an EU or UK developer, the Computer Use lockout alone kills the main reason to move. If you are on a Mac in the US doing web/backend work, the features Codex gained today are nice but do not meaningfully change your day-to-day workflow. If you are a native iOS or macOS developer, Computer Use genuinely opens up workflows that were not practical before and Codex deserves a test.

What today's announcement actually signals is bigger than one product update. OpenAI is done competing on just model quality. They are building Codex into a desktop AI superapp that wraps your whole workflow. Anthropic is still competing on model intelligence and terminal-native agentic coding. Both bets are reasonable. Neither is wrong.

The boring truth: try Codex for a week if you have the budget. Keep Claude Code running alongside it. After seven days, you will know which one fits your actual workflow, not the demo workflow.

Opus 4.7 and Codex Computer Use launching the same day is the strongest signal yet that the AI coding war is fully on. The winners from here are not the subscribers picking the right horse. The winners are the indie hackers shipping faster with whichever tool fits best.

Best Sentry Alternatives for Indie Hackers in 2026 (Honest Picks)

DevToolsPicks — Fri, 17 Apr 2026 05:18:47 +0000

Two things are pushing developers away from Sentry right now.

The first is pricing. Sentry's Developer plan is free but caps at 5,000 errors per month with 1 user. Once you have a real product in production, you blow through that in days. The Team plan starts at $26/month (annual) for 50,000 errors, but overages hit at roughly $0.00029 per error. One bad deploy can generate a surprise bill.

The second is Highlight.io. LaunchDarkly acquired them in April 2025 and shut down the standalone service on February 28, 2026. If you were using Highlight for open-source, self-hostable error tracking, that option is gone. The migration path is to LaunchDarkly Observability, which is a different product with different pricing.

If either of these hit you, here are five honest alternatives for indie hackers and solo developers who just want to know when their app breaks and why.

Quick Verdict

Tool	Best For	Starting Price	Rating
GlitchTip	Drop-in Sentry replacement, self-hostable	Free (self-host) / $15/mo hosted	9/10
Bugsink	Single-container self-hosting, zero maintenance	Free (self-host) / €9/mo hosted	8.5/10
Honeybadger	Small B2B SaaS, bundled with uptime + status page	Free / $26/mo	9/10
Rollbar	Cheapest managed paid tier	Free / $15.83/mo	8/10
Bugsnag	Mobile-first (iOS, Android, React Native)	Free / $23/mo	8/10

The Problem With Sentry at Scale

Before the alternatives, it helps to know exactly why people leave.

Sentry's pricing is event-based. You get 5,000 errors on the free Developer plan, 50,000 on Team at $26/month annual ($29 monthly), and 50,000 on Business at $80/month annual ($89 monthly). Both Team and Business start at the same error quota. The price difference buys you SSO, audit logs, longer retention, and cross-team workflows.

Overages bill at roughly $0.00029 per error event. Sounds tiny. The problem is that one uncaught exception in a loop can generate tens of thousands of events in minutes. A team processing 2 million errors a month can pay around $450/month in error fees alone at published rates, compared to $11/month for 50,000 errors. Session replay, spans, and logs all have their own quotas on top.

Sentry also introduced Seer, their AI debugging agent, at $40 per active contributor per month on top of your base plan. If you have 5 devs contributing, that is $200/month before any Sentry base cost.

None of this makes Sentry a bad product. It is a great product. It is just priced for teams that have VC funding or serious revenue. For indie hackers and solo devs, the numbers get uncomfortable fast.

GlitchTip: The Closest Drop-In Replacement

GlitchTip is an open-source reimplementation of Sentry's backend that uses the same Sentry SDK protocol. That means you can keep your existing @sentry/node, sentry-python, or sentry-php instrumentation, change one URL (your DSN), and GlitchTip receives your errors.

Pricing:

Self-hosted: free, open source, MIT license
Hosted (by Burke Software): free tier for 1,000 events/month, paid plans from $15/month

What it does well:

Sentry SDK compatibility means zero code changes to migrate
Self-hostable with significantly less infrastructure than Sentry (roughly 4 containers vs Sentry's 40+)
EU hosting available, which matters if you have European users and GDPR requirements
Active development and a permissive license

Honest cons:

No session replay, no full distributed tracing, fewer integrations than Sentry
Performance monitoring and uptime features exist but are not under active development
UI is functional but less polished than Sentry
Smaller community, which means slower fixes on edge cases
Still requires Redis, PostgreSQL, Celery, and a frontend/backend split if you self-host

Who should pick GlitchTip: You want to escape Sentry pricing without rewriting instrumentation across your codebase. You are fine giving up session replay and deep tracing. You either want to self-host or pay a small flat fee for managed hosting.

Bugsink: Single-Container Self-Hosting

Bugsink is a fresh implementation of error tracking built specifically for self-hosting. Not a Sentry fork. Not a GlitchTip fork. A clean Django app that runs as a single container. The creator built it after running GlitchTip in production and deciding the multi-service setup was more maintenance than it was worth.

It is also Sentry SDK compatible, so migration is the same one-URL change.

Pricing:

Self-hosted: free, source-available under Polyform Shield license
Hosted (EU-based): flat pricing, not per-event (starts around €9/month for small projects)

What it does well:

Literally one Docker command to run. No Redis, no Celery, no queue
Start with SQLite, switch to PostgreSQL or MySQL for production
Stack traces show source code and local variables inline, which Sentry and GlitchTip both hide by default
The author has been vocal about the Highlight.io shutdown and actively positioning Bugsink as the migration target
Polyform Shield license means you can run it in production and modify it, you just cannot resell it as a competing service

Honest cons:

Deliberately does not support APM, performance monitoring, session replay, or uptime checks
Smaller community than GlitchTip or Sentry
Source-available, not fully open source (the Polyform Shield license restricts commercial resale)
Less mature than Sentry for enterprise features like SAML SSO

Who should pick Bugsink: You want self-hosted error tracking that you set up in 60 seconds and forget about. You do not need APM or session replay. You want to run error tracking on a €5 VPS instead of a Kubernetes cluster.

Honeybadger: The Bundled Value Pick

Honeybadger has been around since 2012. Bootstrapped, developer-led, not chasing VC growth targets. They bundle error tracking, uptime monitoring, cron job monitoring, and a public status page into one subscription. For small B2B SaaS, this replaces 3 or 4 separate tools.

Pricing:

Developer: free for low-volume projects
Small: $26/month for 150,000 errors, unlimited users, uptime + cron monitoring included
Medium: $89/month for 1M errors
Large: $249/month for 5M errors

What it does well:

Uptime, cron job, and status page monitoring all included at the base price
If you exceed your error limit, they process up to 125% before cutting off, and will often absorb one-off overages if you email support
Unlimited users and projects on every paid plan
Strong developer-led support, which means the team that built the product answers your email
Supports Ruby, PHP, Node, JS, Go, Elixir, Python, Java, and more
Integrates with Laravel Forge, which matters if you deploy Laravel apps (we compared Laravel deployment tools here)

Honest cons:

No session replay, which matters for frontend-heavy apps
APM is less detailed than Sentry's performance monitoring
Does not support self-hosting unless you negotiate an enterprise plan
No free tier for teams beyond the solo Developer plan
150K events on the $26/month plan is less than Sentry's 50K. Wait, that is more. Honeybadger gives you 3x Sentry's error quota at the same price point

Who should pick Honeybadger: You run a B2B SaaS with real users, you already pay for Sentry plus uptime monitoring plus a status page. You want to consolidate to one bill. You care more about reliability of alerts than having every observability feature under the sun.

Rollbar: The Cheapest Managed Paid Tier

Rollbar sits in the middle of the market. Not as generous as Honeybadger on bundled features, not as cheap as GlitchTip self-hosted, but the entry paid tier is the cheapest of any managed option here.

Pricing:

Free: 5,000 events per month, 30-day retention
Essentials: $15.83/month for 25,000 events with 90-day retention
Advanced: $32.15/month adds adaptive alerts and extended retention up to 180 days
Enterprise: custom

What it does well:

Machine-learning-based error grouping that reduces noise
Three overage strategies: stop at limit, on-demand continuation, or budget cap
Good CI/CD integration and release tracking
AI-assisted grouping keeps alert fatigue lower than raw Sentry

Honest cons:

Free tier matches Sentry's 5K limit exactly, so no improvement there
Essentials at 25K events is lower than Sentry Team's 50K at a similar price point
No session replay or deep performance monitoring
UI feels dated compared to newer tools
Notification customization is limited

Who should pick Rollbar: You want managed error tracking for under $20/month and your event volume is stable under 25K/month. You do not need session replay or bundled uptime monitoring.

Bugsnag: The Mobile-First Pick

Bugsnag was acquired by SmartBear in 2021 and rebranded as Insight Hub, though most developers still call it Bugsnag. It has always leaned mobile, and the iOS, Android, and React Native SDKs are arguably stronger than Sentry's for crash reporting.

Pricing:

Free: 7,500 events/month
Select: from $23/month for small teams
Preferred: from $39/month for larger teams
Enterprise: custom

What it does well:

The "stability score" metric is the percentage of crash-free user sessions, which is more actionable for mobile release decisions than raw error counts
Stronger mobile SDK coverage than any other tool on this list
Minidump support for native crashes (Electron, Breakpad, Crashpad)
Mature product, owned by a profitable parent company, so stability risk is low

Honest cons:

Cloud-only, no self-hosting option for self-serve plans
No distributed tracing, performance monitoring is an add-on
UI feels dated compared to newer tools
Minidump events count as 5 events each for billing, which makes native app costs unpredictable
The SmartBear acquisition has slowed new feature shipping compared to independent competitors

Who should pick Bugsnag: You build mobile apps. You need crash reporting that understands iOS, Android, and React Native deeply. You care about a stability score you can show leadership or investors as a single release health metric.

How to Choose

Your choice depends on three questions.

Can you self-host? If yes, GlitchTip or Bugsink are effectively free forever, and both use the Sentry SDK so migration is a one-URL change. Bugsink is easier to run (single container), GlitchTip has a larger community.

How much do you care about observability beyond errors? If you want uptime monitoring, cron job monitoring, and a status page bundled in, Honeybadger is the best value. If you only want errors, Rollbar or Bugsnag at their entry tiers are cheaper.

Is your app mobile or web? Mobile-first teams should pick Bugsnag almost regardless of other factors. Web-first teams have more options.

If you are migrating off Highlight.io specifically and you were mainly using it for error tracking (not session replay or the full observability suite), Bugsink is the most frictionless move. Same self-hostable philosophy, Sentry SDK compatible, single-container install.

If you are leaving Sentry because of overage surprises, Honeybadger's "email us if you accidentally exceed limits" policy is genuinely unusual in this category. Most tools either stop processing or bill you.

FAQ

Does GlitchTip really work as a drop-in Sentry replacement?

For error tracking, yes. You change your DSN URL and your existing Sentry SDK instrumentation (@sentry/node, sentry-python, etc.) keeps working. The catch is GlitchTip does not replicate Sentry's full feature set. No session replay, limited distributed tracing, fewer integrations. For pure error tracking, it is a clean swap.

What happens to my Sentry historical data if I migrate?

Most tools will not import historical events from Sentry. You either accept the gap or run both tools in parallel for a transition period. Source maps, alert rules, and issue ownership history also do not migrate automatically. Budget a couple of days for any non-trivial migration just to get source maps working correctly in the new tool.

Is self-hosting Sentry itself a viable alternative?

Technically yes, practically no. Sentry's self-hosted version requires roughly 40 containers, 16GB RAM baseline, and active maintenance for upgrades. The Sentry team does not recommend it for production use. If you want to self-host, GlitchTip or Bugsink are significantly more realistic.

Will my team notice the switch from Sentry?

If you pick a Sentry-SDK-compatible tool (GlitchTip, Bugsink, Better Stack), they will barely notice. Dashboards are different, but alerts still fire, stack traces still show up, and Slack/Discord integrations still work after reconfiguration. If you pick Honeybadger, Rollbar, or Bugsnag, you will need to reinstrument with their SDKs, which is more work.

What about PostHog for error tracking?

PostHog added error tracking in 2024 and has the most generous free tier at 100,000 errors per month. If you already use PostHog for product analytics, consolidating makes sense. If you do not, PostHog is more than you need just for errors. We broke down PostHog against other analytics tools in PostHog vs Plausible vs Fathom vs Mixpanel.

I still need session replay. What should I do?

Of the tools on this list, only Sentry itself has strong session replay. If you need session replay specifically, you are comparing Sentry to PostHog (bundled with analytics) or Hotjar/LogRocket (dedicated). We covered that in Hotjar vs Microsoft Clarity vs PostHog.

Final Recommendation

If you are migrating off Highlight.io: Bugsink. The creator has been the most active voice in the Highlight.io shutdown migration conversation, the architecture is the closest match to what Highlight promised (self-hostable, low-maintenance), and setup takes a minute.

If you want to leave Sentry but keep your instrumentation: GlitchTip. Self-host it for free, or pay $15/month for managed hosting. Sentry SDK compatibility means minimal migration pain.

If you run a B2B SaaS and want to consolidate tools: Honeybadger. Error tracking, uptime, cron monitoring, and status page for $26/month is hard to beat if you were already paying for those things separately.

If you just want the cheapest managed error tracking: Rollbar. $15.83/month for 25K events is the lowest entry paid tier among reputable managed tools.

If your app is mobile: Bugsnag. The mobile SDKs and stability score metric are not matched by anything else on this list.

For most web-first indie hackers, the shortlist is GlitchTip (if you can self-host) or Honeybadger (if you want bundled features and managed hosting). Both are solid picks with different tradeoffs. Either beats staying on Sentry and watching your bill climb every month.

Claude Opus 4.7 Just Launched: What Changed and Is It Actually Worth It?

DevToolsPicks — Thu, 16 Apr 2026 16:13:41 +0000

Anthropic launched Claude Opus 4.7. It is live in Claude.ai for Pro, Max, Team, and Enterprise plans, on the API as claude-opus-4-7, and on AWS Bedrock, Google Vertex AI, and Microsoft Foundry.

Short verdict upfront: if you are on Claude Max and using Opus for coding, switch today. The SWE-bench Pro numbers are a real jump, not marketing. If you are on GPT-5.4 or Gemini 3.1 Pro and happy with your workflow, there is no emergency. Opus 4.7 wins on coding benchmarks but trades blows elsewhere.

The one thing nobody is talking about: the new tokenizer can use up to 35% more tokens for the same text. That matters if you pay per token on the API.

Here is the full breakdown.

Quick Verdict

Model	Best For	Price (input/output per 1M tokens)	Rating
Claude Opus 4.7	Coding, long agentic tasks, vision work	$5 / $25	9.5/10
Claude Opus 4.6	Still fine if you are not pushing limits	$5 / $25	8.5/10
GPT-5.4	General reasoning, some coding tasks	$5 / $25 (similar range)	9/10
Gemini 3.1 Pro	Multilingual, long-context reasoning	Varies by tier	8.5/10

The price on Opus 4.7 is the same as Opus 4.6. That part matters. Anthropic did not charge more for the upgrade, which is unusual for a flagship jump.

What Actually Changed in Opus 4.7

The benchmark chart Anthropic posted tells most of the story. Opus 4.7 jumps meaningfully over Opus 4.6 on every coding and agentic benchmark, and beats GPT-5.4 and Gemini 3.1 Pro on the majority of them.

Coding benchmarks

Benchmark	Opus 4.7	Opus 4.6	GPT-5.4	Gemini 3.1 Pro	Mythos Preview
SWE-bench Pro (agentic coding)	64.3%	53.4%	57.7%	54.2%	77.8%
SWE-bench Verified	87.6%	80.8%	n/a	80.6%	93.9%
Terminal-Bench 2.0	69.4%	65.4%	75.1% (self-reported harness)	68.5%	82.0%

The SWE-bench Pro jump from 53.4% to 64.3% is the headline. That is roughly a 20% relative improvement on the benchmark Anthropic cares about most. For agentic coding work, this is the biggest single-release jump since Opus 4.5.

Terminal-Bench 2.0 is the one place where Opus 4.7 does not lead outright. GPT-5.4 scores higher at 75.1%, but that number uses OpenAI's own harness, not the standard one. Apples to oranges, treat with caution.

Reasoning and other benchmarks

Humanity's Last Exam (multidisciplinary): 46.9% no tools, 54.7% with tools
Agentic search (BrowseComp): 79.3%
Scaled tool use (MCP-Atlas): 77.3%
Agentic computer use (OSWorld Verified): 78.0%
Agentic financial analysis (Finance Agent v1.1): 64.4%
Cybersecurity vulnerability reproduction (CyberGym): 73.1%
Graduate-level reasoning (GPQA Diamond): 94.2%
Visual reasoning (CharXiv Reasoning): 91.0%
Multilingual Q&A (MMMLU): 91.5%

Opus 4.7 leads on most of these. Gemini 3.1 Pro edges ahead on agentic search (85.9%) and ties on some multilingual tasks. Everywhere else, Opus 4.7 is in front.

Vision is the sleeper upgrade

This is the underrated part. Anthropic says Opus 4.7 sees images at more than three times the resolution of Opus 4.6. XBOW's visual-acuity benchmark (used for autonomous penetration testing) jumped from 54.5% on Opus 4.6 to 98.5% on Opus 4.7. That is not incremental. That is a different model for any workflow involving screenshots, UI inspection, PDF processing, or visual document work.

If you are building an app that reads invoices, parses dashboards, or does computer-use automation, this one change may matter more than all the coding numbers combined.

New features in the API

Opus 4.7 ships with three new developer-facing controls worth knowing about:

xhigh effort level. Slots between high and max in the effort parameter. Finer control over the reasoning-vs-latency tradeoff. Use max when you need the absolute best result and do not care about speed. Use xhigh when you need almost that but do not want the latency tax.
Task budgets (beta). Lets Claude prioritize work and manage costs across long runs. Useful for autonomous agents where you would otherwise burn through tokens on a loop.
Adaptive thinking is the default. thinking: {type: "enabled"} with budget_tokens is now deprecated on Opus 4.7. Use thinking: {type: "adaptive"} and let Claude decide how much to think per query.

New features in Claude Code

/ultrareview command. Runs a dedicated review session that flags issues a careful human reviewer would catch. This sits on top of the AI Code Review feature Anthropic shipped earlier this year.
Auto mode for Max users. Reduces interruptions on longer tasks. You set it once, Claude keeps going.
Routines (shipped two days earlier, pairs well with 4.7). Scheduled, API, and GitHub-triggered Claude Code sessions that run on Anthropic's cloud infrastructure. Hand off a bug fix at 2am, come back to a draft PR.

We covered Routines and the desktop redesign in depth in the Claude Code desktop redesign post. If you are trying to make sense of how Skills, MCP connectors, and plugins fit around Opus 4.7, we broke that down in Claude Skills vs MCP Connectors vs Plugins.

The Tokenizer Change Nobody Is Mentioning

Buried in the Opus 4.7 API docs: "Opus 4.7 uses a new tokenizer compared to previous models, contributing to its improved performance on a wide range of tasks. This new tokenizer may use up to 35% more tokens for the same fixed text."

Read that again. Same text, up to 35% more tokens billed.

Opus 4.7 and Opus 4.6 both cost $5 input and $25 output per million tokens. On paper that looks identical. In practice, if your workload lands in the "up to 35% more tokens" range, your real cost per task can rise by roughly a third on the same prompt.

Anthropic says the new tokenizer contributes to the performance improvement, which is believable. More tokens means the model can represent finer-grained concepts internally. But as a developer deciding whether to switch your production workload, you need to test this on your actual prompts before flipping the switch. Do not assume the price tag is the whole price.

For interactive use in Claude.ai on Pro or Max, this does not affect you directly. Your subscription covers it. For API usage in production, benchmark your real workload on both models for a week and compare the actual invoice.

Opus 4.7 vs GPT-5.4

On coding, Opus 4.7 wins. SWE-bench Pro 64.3% to 57.7%, and the Verified number was not published for GPT-5.4 in the Anthropic chart (usually a sign the competitor did not score well).

On general reasoning, it is closer. GPT-5.4 holds up on multi-step agent benchmarks and outscores Opus on self-reported Terminal-Bench. But you have to trust OpenAI's harness for that one.

Honest take: if your primary use case is shipping code, ship on Opus 4.7 or Claude Code. If you are doing general research, writing, brainstorming, or mixed workflows, the choice between GPT-5.4 and Opus 4.7 still comes down to what you are already paying for and which tool you prefer. Neither is definitively better for everything.

We wrote a full breakdown of this tradeoff in ChatGPT Pro $100 vs Claude Max vs Cursor last week. Opus 4.7 strengthens the Claude Max case but does not flip it for everyone.

Opus 4.7 vs Gemini 3.1 Pro

Gemini 3.1 Pro is in a weird spot. It holds its own on multilingual benchmarks and has a strong long-context story. But on the coding work that Claude has historically led in, Opus 4.7 now extends the gap rather than closing it.

If you are already building on Vertex AI and using Gemini for cost reasons, Opus 4.7 is also available on Vertex AI. You can test it side by side in the same console without moving clouds.

If you are building multilingual apps or doing research work where Gemini's long-context handling has served you well, there is no urgent reason to switch. Opus 4.7 is better on most benchmarks, but "better on most benchmarks" and "worth a migration" are different statements.

Who Should Switch to Opus 4.7 Right Now

Switch today:

You are on Claude Max and use Opus heavily for coding. No reason to wait. It is in the model selector. Pick it.
You build agentic workflows with multi-step plans. The SWE-bench Pro jump and the "verifies its own outputs before reporting back" behavior are real. Longer-running tasks are where 4.7 pulls away from 4.6.
You process images, screenshots, PDFs, or do computer-use automation. The 3x resolution vision upgrade is the biggest sleeper change in this release.
You use Claude Code and pay for Max. /ultrareview and Auto mode land today. Worth setting up.

Test before switching:

API users in production. The new tokenizer can eat 35% more tokens. Benchmark on your real prompts first. If your workload is already expensive on Opus 4.6, that 35% could bite.
Teams on Opus 4.6 with stable pipelines. If your prompts are tuned and your outputs are consistent, verify that 4.7 produces the same quality before swapping the model string in production.

Do not bother yet:

You are on Sonnet 4.6 and happy. Sonnet 4.6 is still the price-performance sweet spot for most indie hacker workloads. Opus 4.7 is the flagship, not the default.
You are on the free tier of Claude.ai. Opus 4.7 is Pro, Max, Team, and Enterprise only. Sonnet is what you get on free.
You are on GPT-5.4 for non-coding tasks. If ChatGPT is working for your research, writing, or daily driver use, nothing in this release forces a switch. Revisit when Sonnet 4.8 lands (likely May based on Anthropic's release pattern).

Honest Cons

This is where the DevToolPicks treatment differs from the release-day hype. Every tool has downsides, Opus 4.7 included.

The tokenizer tax. Same $5/$25 rate, but up to 35% more tokens on the same text. For API users, this is a real cost increase disguised as a price match.
Opus is still expensive relative to Sonnet. Most indie hacker workloads should probably still run on Sonnet 4.6 for cost reasons. Opus is the tool you reach for when the task actually requires it.
Mythos overshadows it. Anthropic's benchmark chart includes Mythos Preview, which beats Opus 4.7 on almost every benchmark. Mythos is not publicly available. Seeing a model that is meaningfully better than 4.7 sitting in a closed preview makes 4.7 feel like a midpoint rather than a destination.
Benchmarks are not your workload. SWE-bench Pro is a standardized benchmark. Your codebase is not. The 20% relative jump on SWE-bench Pro does not guarantee a 20% improvement on your actual bugs. Test before you commit.
Prefilling and some older API patterns are deprecated. If you have code built on budget_tokens or thinking: {type: "enabled"}, you need to migrate to adaptive thinking. Small change but not zero.

Pricing and How to Access It

Claude.ai (interactive)

Included in your existing plan:

Pro ($20/month): Opus 4.7 available in the model selector with standard limits
Max ($100-200/month): Higher limits, /ultrareview, Auto mode
Team: Same Opus 4.7 access plus admin features
Enterprise: Same plus compliance and data residency

Free tier does not include Opus 4.7. Sonnet only.

API (`claude-opus-4-7`)

Input: $5 per 1M tokens
Output: $25 per 1M tokens
Prompt caching: Up to 90% cost savings on cache reads
Batch processing: 50% discount on both input and output
Context window: 1M tokens at standard pricing
US-only inference: 1.1x multiplier via the inference_geo parameter

Cloud platforms

AWS Bedrock: Live in us-east-1 and us-west-2. Model ID anthropic.claude-opus-4-7.
Google Vertex AI: Available via base_model: anthropic-claude-opus-4-7. Regional and multi-region endpoints include a 10% premium.
Microsoft Foundry: Available per Anthropic's announcement.

FAQ

Is Opus 4.7 faster than Opus 4.6?

Not necessarily. Opus 4.7 is more thorough and catches more issues during planning. Anthropic says this accelerates overall execution on complex tasks because there are fewer mistakes to fix, but for simple one-shot queries you may see similar or slightly longer response times. If you need raw speed, Opus 4.6 fast mode ($30/$150 per MTok) still exists.

Should I upgrade if I am already paying for Claude Max?

Yes, immediately. It is included. Open the model selector, pick Opus 4.7, done. There is no reason to stay on 4.6 for interactive use if you are on Max.

Is Opus 4.7 better than GPT-5.4 overall?

Better on coding benchmarks. Trades blows on general reasoning. Nobody is definitively better across every task yet. If you write code for a living, Opus 4.7 is the stronger pick today. If you use AI for research or writing, GPT-5.4 is still competitive.

What is Claude Mythos and why is it in the benchmark chart?

Mythos is Anthropic's more powerful unreleased model, held back from public release due to safety and cybersecurity concerns. It is in the chart because it scored higher than Opus 4.7 on several benchmarks, signaling that Anthropic has more capability internally than they are shipping. For now, Opus 4.7 is what you can actually use.

Will my existing Opus 4.6 API code work with 4.7?

Mostly yes. Change claude-opus-4-6 to claude-opus-4-7 and you are most of the way there. Two things to update: migrate from thinking: {type: "enabled"} with budget_tokens to thinking: {type: "adaptive"} with the effort parameter. Prefilling assistant messages is also not supported and will return a 400 error.

When will Sonnet 4.7 launch?

Anthropic's release pattern has Sonnet following Opus by one to four weeks. Internal references to Sonnet 4.8 (not 4.7) have appeared in leaked Claude Code source code, suggesting the next Sonnet release may skip the .7 version entirely and land as 4.8 in May 2026. No official announcement yet.

Final Recommendation

If you have Claude Max: switch to Opus 4.7 in the model selector right now. It is better for coding, better for agentic work, and included in what you are already paying for.

If you use the API in production: test Opus 4.7 on your real prompts for a week before swapping. The tokenizer change is the only part of this release that might cost you money you did not expect. Same price per token, but potentially up to 35% more tokens per task.

If you are on Pro: same answer as Max. Upgrade in the model selector, it is there.

If you are on free: your path to Opus 4.7 is upgrading to Pro at $20/month. Whether that is worth it depends on how much you already use Claude. If you are hitting Sonnet limits weekly, yes. If Sonnet handles everything you throw at it, stay free.

If you are on GPT-5.4 and happy: Opus 4.7 is better for coding on paper, but switching your entire workflow for a benchmark jump is rarely worth it. Wait for Sonnet 4.8 in May, compare price-performance then, and decide.

The real story of this release is not Opus 4.7 beating GPT-5.4 by a few percentage points on SWE-bench. It is that Anthropic shipped a meaningful coding upgrade at the same price as the previous model, in the same week as Claude Code Routines, and with Mythos sitting unreleased in the wings. That is the pattern to watch. Opus 4.7 is where it sits today.

Claude Skills vs MCP Connectors vs Plugins: What Is the Difference and Which Should You Use?

DevToolsPicks — Thu, 16 Apr 2026 13:39:26 +0000

If you have spent any time with Claude recently, you have probably run into these three terms and wondered what makes them different. Skills, MCP Connectors, Plugins. They sound related, they overlap in confusing ways, and Anthropic has not published a single clear diagram showing how they fit together.

This post is that diagram, in words. By the end you will know exactly what each one does, when to use it, and which one you actually need right now.

Quick Verdict

Extension	What It Does	Best For
Skills	Teaches Claude how to do something	Repeatable workflows, custom instructions
MCP Connectors	Connects Claude to external services	GitHub, Google Drive, databases, APIs
Plugins	Bundles Skills + MCP + commands into one install	Team distribution, job-specific toolkits

The Simple Mental Model

Before going deeper, here is the framing that makes everything click:

MCP is the plumbing. It connects Claude to the outside world.

Skills are the instructions. They teach Claude how to do specific things.

Plugins are the finished product. They bundle both together into something you install once.

They are not competing options. They are three layers of the same stack. A Plugin can contain Skills. A Skill can use MCP tools. Understanding which problem you are solving tells you which layer you need.

Read it left to right: a Plugin bundles Skills, MCP Connectors, and Slash Commands together. Skills and MCP both feed Claude. MCP is the only layer that reaches out to external services.

Claude Skills

A Skill is a plain Markdown file (called SKILL.md) that gives Claude specific instructions, context, or a repeatable workflow for a particular task.

Think of it as onboarding documentation for Claude. Instead of re-explaining your preferences, your code conventions, or your review process every single chat, you write it once in a Skill and Claude loads it automatically when relevant.

What Skills are good at:

Teaching Claude your team's coding conventions
Defining a repeatable workflow (deploy process, PR review checklist, blog post format)
Giving Claude domain knowledge it would not have by default
Keeping your main context window clean (Skills only load when needed, costing roughly 30-50 tokens each)

A real example: You review pull requests every week and always want Claude to check the same things: naming conventions, missing tests, security issues. Instead of typing that checklist into every new chat, you create a Skill called pr-reviewer with those instructions. Claude picks it up automatically whenever you ask it to review code.

The honest con: Skills are instructions only. They cannot fetch data, call APIs, or connect to external services. If you need Claude to actually read your GitHub issues or check your database, a Skill cannot do that on its own. That is what MCP is for.

Who should NOT use Skills: If your problem is connecting Claude to an external service, Skills are the wrong tool. They shape behavior, not connectivity.

MCP Connectors

MCP stands for Model Context Protocol. It is an open standard (now under the Linux Foundation, co-founded with Block and OpenAI) that defines how AI models connect to external tools and data sources.

The USB-C analogy keeps coming up for good reason: build an MCP server once and it works with Claude, ChatGPT, Gemini, or any other tool that speaks the protocol. OpenAI adopted it in March 2025. Google DeepMind followed in April.

In the Claude context, Connectors are the pre-built MCP integrations Anthropic and third parties have packaged for common services: Microsoft 365, Google Drive, Slack, GitHub, Linear, and more. When you enable a Connector, Claude can actually read your emails, access your files, and interact with those services on your behalf.

What Connectors are good at:

Giving Claude real-time access to external data (emails, documents, repositories)
Acting as you with delegated permissions via OAuth (if you cannot see a SharePoint doc, Claude cannot either)
Connecting Claude to your actual workflow without writing code

The honest con: MCP Connectors consume a significant amount of context before you type a single word. A five-server MCP setup can use around 55,000 tokens of context overhead upfront. That is a third of your context window gone before the conversation starts. Install only what you actually use.

Who should NOT use Connectors: If you just want to teach Claude how to follow your workflow or coding style, you do not need MCP. That is a Skill.

Plugins

Plugins are the newest addition, shipping with Claude Cowork on January 30, 2026. They landed hard enough to wipe $285 billion off software stocks on the day of announcement.

A Plugin bundles Skills, MCP Connectors, slash commands, and sub-agents into a single installable package. If Skills are individual recipe cards and MCP is the kitchen plumbing, a Plugin is the full kitchen, stocked and ready to cook.

Anthropic launched 11 official Plugins at release, covering legal, sales, finance, marketing, data analysis, and more. A Plugin marketplace launched in February 2026.

What Plugins are good at:

Distributing a complete workflow to a team without requiring anyone to configure things manually
Packaging a job-specific toolkit (a legal Plugin includes NDA review workflows, document management connectors, and slash commands like /review-contract)
Ensuring everyone on a team has the same setup with one install

A real example: A sales team installs a Plugin that connects to their CRM via MCP, includes Skills for their prospecting workflow, and exposes commands for research and follow-up drafts. New hire joins, installs the Plugin, and Claude already knows how the team works.

The honest con: Plugins are overkill for individual use. If you are a solo developer or indie hacker, you do not need to package and distribute anything. Just write Skills directly for your own workflows and add MCP servers for external services you actually use.

Who should NOT use Plugins: Solo developers working alone. The overhead of building a Plugin is only worth it when you need to distribute a setup to others.

How They Fit Together in Practice

These three layers are not mutually exclusive. Real setups combine them:

A developer building a SaaS might have:

A Skill that teaches Claude their codebase conventions and PR review checklist
An MCP Connector for GitHub so Claude can read actual issues and open PRs
No Plugin needed, because they are working alone

A small team might have:

A Plugin that packages their deployment workflow Skill, a GitHub MCP Connector, and a /deploy slash command
One install and the whole team is set up identically

The decision is simpler than it looks. Ask yourself three questions:

Do I need Claude to connect to an external service? Yes: use an MCP Connector. No: skip it.
Do I have a repeatable workflow or specific instructions I keep re-explaining? Yes: write a Skill. No: skip it.
Do I need to distribute this setup to others? Yes: package it as a Plugin. No: skip it.

Which One Do You Actually Need Right Now?

You are a solo developer or indie hacker using Claude for coding:
Start with Skills. Write a SKILL.md with your project conventions, preferred patterns, and any workflows you repeat. Add a GitHub MCP Connector if you want Claude to actually interact with your repository. You do not need a Plugin.

You are using Claude for content or writing:
Skills are your best tool. Define your voice, your post structure, your review checklist. Claude will apply it consistently without you repeating yourself.

You want Claude to read your emails, files, or documents:
MCP Connectors are what you need. Enable the Google Drive or Microsoft 365 connector and Claude can access your actual files.

You are setting up Claude for a team:
Build a Plugin. Package your Skills, Connectors, and any slash commands into one install so everyone gets the same setup without manual configuration.

FAQ

Do Skills work in both Claude.ai and Claude Code?

Yes. The same SKILL.md format works across Claude.ai, Claude Code, and the API. Skills have been adopted as an open standard and also work with Codex and Gemini CLI.

Do I need to code to use MCP Connectors?

No. Pre-built Connectors for common services like Google Drive, Microsoft 365, and GitHub are available through Claude's settings without writing any code. Custom MCP servers require development work.

Are Plugins free?

The Plugin marketplace launched in February 2026. Anthropic's official Plugins are free and open-source on GitHub. Third-party Plugins vary by provider.

What is the difference between Skills in Claude.ai and CLAUDE.md in Claude Code?

They serve the same purpose but in different contexts. CLAUDE.md is for Claude Code (the terminal tool) and applies to a specific project. Skills in Claude.ai work across your conversations. Both give Claude persistent instructions without requiring you to repeat yourself every session.

Should I be worried about security with third-party Skills?

Yes, with caution. Snyk's ToxicSkills research from February 2026 found that 13.4% of publicly available Skills had critical vulnerabilities. Skills can include executable scripts. Read any SKILL.md file you install from an external source before enabling it, the same way you would review an npm package.

The Bottom Line

The confusion is understandable. These features rolled out at different times, work slightly differently across Claude.ai and Claude Code, and the names do not make the hierarchy obvious.

But the mental model is clean once you have it:

Skills teach Claude how to work. Write them for any workflow you repeat more than twice.

MCP Connectors give Claude access to the outside world. Install them when you need Claude to read your actual data.

Plugins distribute everything as a kit. Build them when you are setting up multiple people.

Most solo developers only need Skills and one or two Connectors. Plugins come later, if at all.

If you are evaluating Claude as part of a broader AI tool decision, the ChatGPT Pro vs Claude Max comparison and the Zapier vs Make vs n8n automation breakdown cover how Claude fits into the wider solo developer stack.

Cal.com vs Calendly vs TidyCal for Indie Hackers in 2026: Which Scheduling Tool Is Worth It?

DevToolsPicks — Thu, 16 Apr 2026 05:45:15 +0000

Every indie hacker eventually needs a booking link. A potential user wants to jump on a call. A client wants a demo. A collaborator needs 30 minutes. And then you realise your options are: Calendly (everyone knows it, costs money to unlock basics), Cal.com (open-source, generous free tier, beloved by developers), or TidyCal ($29 once, no subscription, done).

This post is a direct comparison of all three for solo founders and indie hackers. Not for enterprise sales teams. Not for recruiting pipelines. For the person building a SaaS, freelancing, or running a one-person operation who needs a clean booking link without overpaying.

Quick Verdict

Tool	Best For	Pricing	Free Plan
Cal.com	Developers, API integrations, self-hosters	Free forever (solo) / $15/user/month (teams)	Yes, and genuinely generous
Calendly	Non-technical users, sales teams	$10/user/month (annual)	Yes, but very limited
TidyCal	Solo founders who want to pay once and forget it	$29 one-time (lifetime)	Yes, basic features only

Cal.com

Cal.com is open-source, self-hostable, and has 40k+ stars on GitHub. It was built because developers got frustrated with Calendly's restrictions and decided to build something better. The result is a scheduling tool that gives you a lot for free and scales without surprising you with a bill.

Pricing:

Free: Unlimited event types, multiple calendar connections, workflow automation, routing forms, payment collection, and webhooks. Single user only with Cal.com branding on booking pages.
Teams: $15/user/month (14-day free trial)
Organizations: $37/user/month (SSO, SCIM, advanced security)
Self-hosted: Free. You pay only for your server. Open-source under AGPL license.

The free plan is genuinely unusual. Most tools lock workflows, automations, and payment collection behind a paywall. Cal.com includes all of that for free as long as you're a solo user. The catch is the Cal.com branding on your booking pages. If that bothers you, you're either upgrading to a paid plan or self-hosting.

The API is open and well-documented. Developers building scheduling into their own products use Cal.com's Platform plan to embed the booking flow. The self-hosting option is real, not theoretical. If you're already running a VPS for your SaaS, hosting Cal.com alongside it is a straightforward afternoon project. The developer experience is strong and the product ships fast.

The interface is more complex than Calendly's. When you first log in, you see a lot of options. For someone who just wants a booking link in five minutes, Cal.com can feel like more product than needed. The Cal.com branding on the free plan is also a meaningful limitation if you're sharing booking pages with clients who'll judge the professionalism of the link.

Skip Cal.com if you're non-technical and want something set up in under 10 minutes without any configuration. Also worth noting: if you need deep Salesforce or HubSpot integration out of the box, Cal.com's integrations ecosystem is thinner than Calendly's in that specific area.

Calendly

Calendly invented the "send-a-link" scheduling experience. It's what most people have seen before. Share a Calendly link with someone outside the tech world and they know exactly what to do. That familiarity is worth something.

Pricing:

Free: 1 event type, 1 calendar connection, unlimited meetings. Very limited in practice.
Standard: $10/user/month (annual billing) or $12/user/month (monthly). Includes unlimited event types, 6 calendar connections, payments, and reminders.
Teams: $16/user/month (annual) or $20/user/month (monthly). Adds round-robin, Salesforce/HubSpot, routing forms, and analytics.
Enterprise: $15,000+/year

The free plan's single-event-type limit is the most common frustration. You set up a 30-minute call link, and then you want to add a 15-minute intro call, and suddenly you need to pay $10/month. Most solo founders hit this ceiling within a week.

Polish is where Calendly earns its reputation. The booking page looks great. The mobile app is strong. Onboarding takes five minutes. If your clients are non-technical and you want them to have a frictionless experience, Calendly delivers that better than anything else here. The integration ecosystem is the deepest of the three: 700+ integrations via Zapier plus native connections to Salesforce, HubSpot, and the usual video call tools.

The per-user billing model gets expensive as you grow. Two users on the Teams plan are paying $32/month annually ($384/year). A five-person team is $80/month. For a solo founder this isn't ruinous, but it's recurring cost for a tool that competes directly with Cal.com's free tier. The free plan is also notably more restricted than Cal.com's. One active event type is tight for anyone managing more than one type of meeting.

If you have any technical ability and want to pay nothing or as little as possible, Cal.com is the better call. The only reason to pay for Calendly's Standard plan as a solo user is if you specifically need branding-free booking pages and polished UI with zero configuration.

TidyCal

TidyCal is an AppSumo Original, built by the AppSumo team and sold primarily as a lifetime deal. The pitch is simple: pay $29 once, get a scheduling tool with no subscription, no renewals, no billing anxiety at the end of every month.

Pricing:

Free: Unlimited bookings, 1 calendar connection, limited integrations
Individual: $29 one-time (lifetime). Includes 10 calendar connections, Zoom, Google Meet, Zapier, PayPal, Stripe, and unlimited booking types.
Agency: $79 one-time (lifetime). Adds team features, round-robin, and collective bookings.
Both plans come with AppSumo's 60-day money-back guarantee.

The math is compelling for anyone used to monthly SaaS costs. Calendly Standard costs $120/year. Over three years that's $360. TidyCal costs $29, total, forever. If the tool does 80% of what you need, the one-time price makes it an obvious call for cost-conscious founders.

The simplicity is genuine. You can have a booking page live in under 15 minutes. Payments through Stripe and PayPal carry zero commission, so you keep everything your clients pay. The unlimited booking types on the paid plan cover most solo founder use cases. Date polls and recurring bookings are also included.

The integration count is the main weakness. TidyCal has 14 native integrations versus Calendly's 700+. There is no mobile app in 2026. No 2-factor authentication. No Microsoft Exchange support (only Microsoft 365 and Outlook.com). The team features require all team members to have purchased their own TidyCal license, which is counterintuitive and buried in the documentation.

The honest comparison worth making: Cal.com's free plan gives you more than TidyCal's paid plan on most feature dimensions. If you're a developer comfortable with a slightly more complex setup, Cal.com free beats TidyCal $29. TidyCal wins on simplicity and the "pay once" comfort, not on raw features.

Skip TidyCal if you need deep integrations with your existing SaaS stack, a mobile app, or team features that work without requiring every team member to buy a separate license.

Head-to-Head Comparison

Free plans

Cal.com's free plan is the most generous by a significant margin. Unlimited event types, multiple calendar connections, automations, payment collection, and routing forms are all included at no cost for a single user. The only real limits are branding and the single-user restriction.

Calendly's free plan gets you one event type. That's it. You'll hit it immediately.

TidyCal's free plan gives you unlimited bookings but only one calendar connection and no Zoom or payment integrations. It's fine for testing the product, not for running a business on it.

Winner: Cal.com

For non-technical users

Calendly wins here cleanly. The interface is the most polished, the setup is the fastest, and the booking page is familiar to recipients. If you're sharing booking links with people who aren't technical and you want zero friction, Calendly is the right call.

TidyCal is a reasonable second option, simple enough for most people. Cal.com has a learning curve. Not steep, but present.

Winner: Calendly

Lifetime cost for a solo founder over 3 years

Cal.com free: $0
TidyCal Individual: $29 total
Calendly Standard: $360 total ($10/month x 36 months)

This isn't even close if you're happy on a free plan. But if you want branding-free booking pages, TidyCal is $29 versus $360 for Calendly Standard.

Winner: Cal.com (free) or TidyCal (paid)

Developer friendliness

Cal.com runs away with this. Open-source codebase, full API on the free plan, webhooks, self-hosting, and a Platform plan for embedding scheduling into your own product. If you want to integrate scheduling into your SaaS as a feature, Cal.com is the tool.

Winner: Cal.com

How to Choose

You're a solo founder or indie hacker with any technical ability: Use Cal.com free. You get more than you'll need for zero cost. The setup takes 20 minutes. The only thing you'll miss is branding-free booking pages.

You want to pay once and stop thinking about it: TidyCal at $29. You get a solid booking tool, Stripe payments, Zoom integration, and unlimited booking types. No monthly guilt. The $29 pays for itself after two or three bookings if you're charging for your time.

You need polish and your clients are non-technical: Calendly Standard at $10/month. It's the most frictionless experience for booking recipients and the 700+ integrations matter if you're already using HubSpot, Salesforce, or complex automation stacks.

You want to build scheduling into your own product: Cal.com Platform plan. Nothing else comes close for developer-grade embeddable scheduling.

FAQ

Is Cal.com actually free forever?

Yes, for individual use. Unlimited event types, multiple calendar connections, automations, and payment collection are all on the free plan. You'll need to upgrade to a paid plan or self-host to remove Cal.com branding from booking pages.

Is TidyCal still being actively developed in 2026?

Yes. AppSumo Originals actively ships updates to TidyCal users, and the lifetime plan includes all future updates at no extra cost. The latest updates added round-robin and collective booking types plus a new booking editor. It's not abandoned software.

Why would anyone choose Calendly over Cal.com?

Mainly for polish and non-technical users. Calendly's booking page is widely recognised and trusted. The mobile app is stronger. The integration ecosystem is deeper for sales-heavy stacks. If your priority is zero friction for recipients and you don't mind paying $10/month, Calendly earns it.

Can I self-host Cal.com on my own VPS?

Yes. Cal.com is open-source under the AGPL license. If you're already running a VPS for your SaaS (and most indie hackers reading this are), hosting Cal.com alongside it is a viable option. The documentation for self-hosting is solid.

Does TidyCal take a cut of paid bookings?

No. TidyCal charges 0% commission on payments processed through Stripe or PayPal. You keep everything your clients pay.

The Bottom Line

For most indie hackers: Cal.com free is the obvious starting point. It's genuinely generous and requires no ongoing cost. The only friction is that it takes slightly more setup time than Calendly.

If you want to pay once and forget it: TidyCal at $29. Not the most feature-rich option, but $29 once beats $120/year on Calendly Standard by a mile over any reasonable time horizon.

If you're non-technical or your clients definitely aren't: Calendly Standard. The polish is real and the $10/month is reasonable.

For a deeper look at the rest of your indie hacker stack, the Zapier vs Make vs n8n comparison and the PostHog vs Plausible vs Fathom analytics breakdown are worth reading alongside this.

Claude Code Just Redesigned Its Desktop App for Parallel Sessions: What Changed and Is It Worth It?

DevToolsPicks — Wed, 15 Apr 2026 05:32:48 +0000

Anthropic shipped a full redesign of the Claude Code desktop app on April 14, 2026. Not a minor update. A from-scratch rebuild of the interface around a different mental model: you're not waiting for one task to finish, you're running several at once and checking in as results arrive.

If you've been using Claude Code as a terminal tool and wondering when the desktop experience would catch up, today is that day.

Here's what actually changed, what it means for solo devs, and the honest limitations.

What Anthropic shipped

Two things landed simultaneously today. The desktop redesign is the one getting attention on X, but there's a second announcement worth knowing about: Routines, scheduled automations that run on Claude Code's cloud infrastructure even when your laptop is off. Think cron jobs, but AI-native. A routine packages a prompt, one or more repos, and a set of connectors, then runs on a schedule or trigger. Research preview for now, but worth watching.

The desktop redesign is what this post covers.

The new sidebar and parallel sessions

The headline feature is the sidebar. Every active and recent session lives there. You can kick off work across multiple repos and move between them as results come in.

You can filter sessions by status, project, or environment. Group the sidebar by project to find things faster. When a session's PR merges or closes, it archives itself automatically so the sidebar stays focused on what's actually live.

The mental model shift here is real. Before this update, Claude Code worked like a single focused conversation. You had one thing in flight. The new app is built for what Anthropic calls "the orchestrator seat": you have three things running (a refactor in one repo, a bug fix in another, a test-writing pass in a third), and you're checking on each as results come in, steering when something drifts, reviewing diffs before you merge.

For a solo founder running Claude Code as their de facto engineering team, this is a significant change. You're not babysitting one task. You're managing several.

Side chat without interrupting your task

The feature I'm most interested in is side chat. Press ⌘ + ; (or Ctrl + ;) to open a branched conversation. Side chats pull context from the main session thread, but nothing you say in side chat gets added back to the main thread.

The practical use case: Claude is halfway through a large refactor. You want to ask it a question about approach without accidentally redirecting the work already in progress. Previously you'd either interrupt the task or open a separate session with no context. Now you get a conversation that knows what's happening without misdirecting it.

Small feature. Real quality-of-life difference.

Integrated terminal, file editor, and faster diffs

The redesign brings three tools into the app that used to require context-switching to your editor:

Integrated terminal. Run tests or builds alongside your session. The biggest pain point with Claude Code previously was having to jump to a separate terminal to verify whether Claude's output actually worked. That context switch is gone.

In-app file editor. Open files, make spot edits, save changes without leaving Claude Code. For minor tweaks this is much faster than the round trip to VS Code.

Faster diff viewer. Rebuilt for performance on large changesets. This one matters if you've run Claude Code on a big refactor and then watched the diff viewer struggle. The rewrite should handle that better.

There's also an expanded preview pane. You can open HTML files or PDFs in-app, plus run local app servers in the preview. Less useful for backend-heavy work, but frontend and full-stack developers will use this constantly.

Every pane is drag-and-drop. Arrange the terminal, preview, diff viewer, and chat in whatever grid matches how you work.

View modes and shortcuts

Three view modes let you control how much you see:

Verbose: full transparency into Claude's tool calls
Normal: balanced view
Summary: just the results

For developers who find AI coding tools overwhelming when they show every micro-decision, Summary mode is a genuine improvement. For those who want to understand exactly what Claude is doing and why, Verbose keeps that available.

New keyboard shortcuts cover session switching, spawning, and navigation. Press ⌘ + / (or Ctrl + /) to see the full list. A new usage button shows both your context window and session usage at a glance, which matters for managing token budgets on heavier workloads.

Plugin parity and SSH for Mac

Two items in the release notes that deserve mention:

Plugin parity. CLI plugins now work in the desktop app exactly the way they do in the terminal. If your organisation manages Claude Code plugins centrally, or you've installed your own locally, they work here now. This closes a real gap.

SSH support extended to Mac. Previously SSH was Linux only. You can now point desktop sessions at remote machines from Mac. For solo devs with a remote dev machine or VPS, this matters.

Who gets access

The redesigned desktop app is available for all Claude Code users on Pro, Max, Team, and Enterprise plans, plus via the Claude API. It's not available on the free plan.

If you're on Pro ($20/month) or Max, update and restart the app. If you don't have the desktop app yet, download it at claude.com/download.

Linux users: the redesigned desktop app is not available for Linux yet. The terminal-based Claude Code CLI still works on Linux, but the new visual experience is macOS and Windows only for now.

How does this compare to Cursor?

Claude Code has been Cursor's main competitor for the serious agentic coding use case. The Cursor 3 launch in April 2026 also brought parallel agents and a multi-session interface. The gap between them was: Cursor had a polished, integrated visual experience; Claude Code had raw capability in a terminal.

Today's release closes that gap significantly. The integrated terminal, file editor, diff viewer, and drag-and-drop layout put Claude Code's desktop experience in the same tier as Cursor's. The key remaining difference is that Cursor is built on a full IDE (VS Code fork), while Claude Code is purpose-built for agentic sessions. That's a philosophical difference as much as a feature one. One is "your IDE with Claude built in." The other is "an AI coding environment that can call your tools."

Neither approach is wrong. They suit different workflows. If you're comparing, this update is worth re-evaluating that decision. You can see a full breakdown in the Claude Code vs Cursor vs GitHub Copilot comparison.

What's still missing

Being honest here:

Linux desktop support. The CLI works on Linux, but the redesigned visual app doesn't. If you're a Linux developer, you're still on the old experience.

Routines is a research preview. The scheduled automation feature sounds powerful but it's explicitly early. Expect rough edges.

Plan requirement. You need a paid plan. The free tier doesn't include Claude Code desktop. If you're evaluating whether the subscription is worth it, the Claude Max vs ChatGPT Pro vs Cursor comparison breaks down the value question.

Is it worth updating?

Yes, immediately, if you're on a paid plan and using Claude Code on macOS or Windows.

The parallel sessions alone change how you use the tool if you're running multiple things at once. Side chat removes one of the more frustrating friction points. The integrated terminal means you can actually verify Claude's work without leaving the app.

This isn't a marketing update. It's a real shift in how the desktop experience works.

If you're not using Claude Code yet and are comparing options, today's release makes it harder to justify dismissing it as "just a terminal tool." It's not that anymore.

FAQ

Does the Claude Code desktop redesign cost extra?

No. It's included with existing Pro, Max, Team, and Enterprise plans. Update and restart the app to get it.

Is the redesigned Claude Code app available on Linux?

Not yet. The CLI still works on Linux. The new visual desktop experience is currently macOS and Windows only.

What is Claude Code Routines?

Routines are scheduled automations that run on Anthropic's cloud infrastructure, even when your laptop is off. They package a prompt, one or more repos, and connectors into a repeatable task. Launched today as a research preview alongside the desktop redesign.

How does Claude Code compare to Cursor after this update?

Both now have parallel session support and integrated visual interfaces. The main difference is architecture: Cursor is a VS Code fork with AI built in; Claude Code is purpose-built for agentic sessions. See the full Cursor vs Windsurf vs Zed comparison for more context on where each editor fits.

What plans include the new Claude Code desktop app?

Pro ($20/month), Max, Team, and Enterprise plans, plus the Claude API. Not available on the free plan.

Best Intercom Alternatives for Indie Hackers in 2026 (Honest Picks)

DevToolsPicks — Tue, 14 Apr 2026 05:22:09 +0000

Intercom is genuinely impressive software. The messenger is beautiful, the automations are solid, and the Fin AI agent does cut down repetitive tickets. But the bill? That's where things get painful for bootstrapped founders.

Here's the math nobody puts in a sales deck: Essential plan is $39/seat/month on monthly billing. Add one Fin AI resolution per conversation, and at 500 resolved tickets a month you're paying another $495. That's for a solo founder with one support seat and moderate traffic. For something under $1,000 MRR, that's not a support tool. That's a competitor.

The good news is 2026 has better options. Not watered-down options. Real ones that handle live chat, shared inboxes, knowledge bases, and AI without treating every resolved ticket like a taxable event.

Here are the five I actually recommend.

Quick Verdict

Tool	Best For	Starting Price	Free Plan
Crisp	Early-stage SaaS founders	€45/month	Yes (2 agents)
Tawk.to	Pre-revenue / zero budget	Free forever	Yes (unlimited)
Help Scout	Email-first, high-touch support	$25/user/month	Yes (5 users)
Chatwoot	Technical founders, full control	Free (self-hosted)	Yes
Tidio	Consumer apps and e-commerce	$29/month	Yes (50 chats)

1. Crisp

Crisp is the one you'll see in almost every early-stage SaaS. It's the most common answer on Indie Hackers and Product Hunt when someone asks "what do you use for support?" There's a reason for that: it works, it's cheap, and it doesn't punish you for growing.

Pricing:

Free: 2 agents, live chat widget, basic inbox
Mini: €45/month (4 agents, email support, basic triggers)
Essentials: €95/month (10 agents, omnichannel inbox, knowledge base, limited AI)
Plus: €295/month (20 agents, full AI, ticketing, white-label)

All plans are flat-rate per workspace, not per conversation. No $0.99 per AI resolution surprises.

The widget is tiny and fast. It won't tank your Lighthouse score. The free tier is generous enough to run support for an early-stage SaaS without feeling like a demo. The Essentials plan is where most indie hackers end up, and for €95/month you get omnichannel inbox, a knowledge base, and enough AI to handle common questions without hiring. The Plus plan is where Crisp starts competing directly with Intercom on features, at a fraction of the price.

The weak spot is the AI cap on Essentials. If you want genuinely unlimited AI resolutions, you need Plus at €295/month, which is a significant jump. The Mini plan also feels underpowered. Most small teams hit its ceiling quickly and end up upgrading to Essentials anyway, so you might as well start there.

Skip Crisp if your customers live in Slack. It's a chat-and-inbox tool, not a Slack-native support hub. For B2B SaaS with enterprise clients expecting Slack Connect threads, look elsewhere.

2. Tawk.to

The most frequently cited answer to "what's actually free?" is tawk.to. And unlike most "free" tools, tawk.to means it. Unlimited agents, unlimited chats, unlimited websites. No seat cap. No conversation limit. The core product is free, permanently.

Pricing:

Free forever: unlimited agents, live chat, ticketing, knowledge base, CRM
Branding removal: $29/month (removes "Powered by Tawk.to" from the widget)
AI Assist: from $29/month (automated responses, 24/7 coverage)
Video + voice + screen sharing: $29/month

The business model is essentially: the free chat tool brings customers in, then optional services like hired agents ($1/hour) and AI add-ons generate revenue. It works because the core software is genuinely funded by those services.

Nothing else in this category is this free with this many features. If you have zero budget and need a live chat widget today, tawk.to is the right call. You get a ticketing system, a knowledge base, a basic CRM, and unlimited agents out of the box. A ten-person indie team could use this indefinitely without paying a cent, as long as the Tawk.to branding on the widget doesn't bother them.

It does have real limitations though. The product isn't as polished as Crisp or Intercom. The mobile app has reliability issues that users mention consistently. Reports and analytics are sparse. The AI Assist feature is a paid add-on, not built in. And if you want a clean, branded widget, you're paying $29/month minimum anyway, which puts you close to Crisp Mini pricing but with less capability.

If you care about aesthetics or want modern automation workflows, tawk.to will feel dated. It's built for function, not design.

3. Help Scout

Help Scout takes a different view of customer support: conversations, not tickets. There's no ticket number in your customer's face. They just get an email reply. It feels like talking to a person, not filing a complaint.

Pricing:

Free: 5 users, 1 inbox, 100 contacts/month
Standard: $25/user/month (unlimited inboxes, automations, live chat via Beacon)
Plus: $45/user/month (advanced reporting, AI Drafts, CRM integrations)
Pro: $65/user/month (SSO, HIPAA compliance, dedicated support)
AI Answers: $0.75/resolution (add-on, not included in base plans)

Help Scout recently shifted to a contacts-based model. You're charged based on the number of customers you help each month, not purely per seat. If the same customer sends you five messages in one month, that counts as one contact. This is more predictable than Intercom's per-resolution model.

The interface is clean and feels like Gmail, which means almost no training required for new team members. For bootstrapped SaaS with customers who prefer email over chat, it's the best option in this list. The Beacon widget adds live chat without much friction. AI Drafts on the Plus plan generates suggested replies based on your knowledge base, which is the kind of AI that actually saves time without charging per resolution.

The per-user pricing is the main drawback. Costs scale with your team. The free plan's 100-contact limit is too tight for anyone with real traffic. Standard gives you limited reporting (2 years of history only), and if you want proper analytics or Salesforce/HubSpot integrations, you're jumping to Plus at $45/user/month. AI Answers is also billed separately on top of base plan prices.

Help Scout isn't the right fit if your users expect instant live chat. It's email-first and that shapes everything about how the product works.

4. Chatwoot

Chatwoot is open-source. That's the headline. You can self-host it on your own VPS for free, with full control over your data and no subscription fee beyond what your server costs. For a developer who already has infrastructure up and running, this is a serious option.

Pricing:

Open-source self-hosted: Free (MIT License), you pay only for hosting
Cloud Free: 2 agents, basic live chat, 30-day data retention
Cloud Startups: $19/agent/month (all channels, no retention limits)
Cloud Enterprise: $99/agent/month (dedicated support, advanced security)

On a $5-$10/month VPS you'd already have running, you can host a full multi-channel support system with no monthly SaaS fee. That's the pitch.

The omnichannel inbox is genuinely strong. Email, WhatsApp, Facebook Messenger, Twitter, and live chat all feed into one place. The API is solid if you want to build custom integrations from your Laravel app. For a technical solo founder who wants to own their support stack completely, there's nothing better at this price point. The Startups cloud plan at $19/agent/month is also competitive if you want someone else handling the hosting.

Self-hosting Chatwoot means you're maintaining it. Updates, backups, infrastructure management are on you. If you don't enjoy that kind of work, the "free" price has a real cost in time. The cloud free plan is limited to a point where it's not useful for production support. The UI is functional but not polished.

If you don't want to manage servers, use the cloud version. If you want sophisticated AI automation out of the box, Chatwoot isn't there yet. It's a solid foundation that requires your own investment to get right.

5. Tidio

Tidio is built for consumer apps and e-commerce more than SaaS. But it earns a spot here because the Lyro AI chatbot is genuinely impressive, the onboarding is fast, and the free plan is workable if your support volume is low.

Pricing:

Free: 50 conversations/month (live chat + basic helpdesk)
Starter: $29/month (100 conversations, live chat)
Growth: $59/month (250+ conversations, removes Tidio branding)
Tidio+: $749/month (scaling teams, dedicated support)

Lyro AI is the reason to use Tidio. It's a conversational AI that pulls answers from your knowledge base and handles common questions without human involvement. Lyro can resolve up to 70% of repetitive support queries automatically, and unlike Intercom's Fin, it doesn't charge per resolution at the lower tiers. The visual chatbot builder is no-code. Setup took under 10 minutes in testing.

The pricing gap between Growth and Tidio+ is a real problem. You go from $59/month to $749/month with nothing in between. A team that outgrows the Growth plan has no graceful upgrade path. The free plan's 50-conversation cap runs out fast for any product with real users. The platform also skews toward e-commerce workflows, and some SaaS-specific features feel bolted on as an afterthought.

If you're building B2B SaaS and your customers want technical support conversations, not automated chat flows, Tidio will feel like the wrong tool. It's optimized for volume, not depth.

How to Choose

If you're pre-revenue or under $500 MRR: Start with Tawk.to. It's free, functional, and doesn't require a credit card. When the Tawk.to branding starts bothering you (and it will), migrate to Crisp's free tier and eventually Mini.

If you're an early-stage SaaS under $5,000 MRR: Crisp Essentials at €95/month is where most indie hackers end up. Flat pricing, omnichannel inbox, enough AI to handle the easy stuff. You won't be surprised by your bill.

If your customers prefer email over chat: Help Scout Standard at $25/user/month. It's the cleanest email-first support tool available and your customers won't even know they're using a helpdesk.

If you want to own your stack completely: Chatwoot self-hosted. If you already have a VPS running your SaaS (which you probably do if you're reading DevToolPicks), hosting Chatwoot alongside it costs almost nothing. Laravel integrates with Chatwoot's API cleanly.

If you're building a consumer product with high chat volume: Tidio's Lyro AI handles repetitive queries better than anything else in this price range. Start with the free plan and move to Starter when you hit the limit.

FAQ

Is there a free Intercom alternative?

Yes. Tawk.to is free forever with unlimited agents and chats. Crisp and Help Scout both have free tiers, though with limits. Chatwoot can be self-hosted for free.

Why is Intercom so expensive for solo founders?

Intercom's pricing has two layers: per-seat costs ($29-$139/seat/month depending on plan) and per-AI-resolution fees ($0.99 per resolution). As your AI handles more tickets, the second number grows regardless of your plan. For low-MRR founders, that combination is hard to justify.

Which Intercom alternative is closest to Intercom in features?

Crisp on the Plus plan is the closest feature match at a lower price. You get an omnichannel inbox, AI chatbot, knowledge base, ticketing, and analytics. The UI is different but the functionality is comparable.

Does Crisp have a free plan?

Yes. Crisp's free plan includes 2 agent seats, a live chat widget, and a shared inbox. It's genuinely usable for a very early-stage product and doesn't expire.

Can I migrate from Intercom without losing conversation history?

It depends on the tool. Crisp, Help Scout, and Chatwoot all have migration guides or CSV import options. Conversation history migration is usually partial at best. Budget for some setup work when switching.

The Bottom Line

Intercom makes sense for funded SaaS companies with a real support team and the volume to justify variable AI pricing. For most indie hackers, that's not where you are yet.

Crisp is the default pick. Free tier gets you started, Essentials at €95/month covers most solo founders for years, and the pricing doesn't scale based on how many tickets your AI resolves.

Tawk.to if you need free today. No strings, no hidden fees, unlimited everything.

Help Scout if your customers are email-first and you want to feel human at scale.

Chatwoot if you want to own your data and are comfortable managing a server.

If you're still figuring out your support stack while building your SaaS, I've got a comparison of analytics tools for indie hackers and a breakdown of the best email marketing tools that pair well with whichever support tool you pick here.

Cursor vs Windsurf vs Zed for Indie Hackers in 2026: Which AI Code Editor Is Actually Worth It?

DevToolsPicks — Mon, 13 Apr 2026 05:35:20 +0000

The AI code editor market has split into three distinct camps. Cursor dominates on ecosystem and community. Windsurf leads on autonomous agent capability. Zed is doing something different entirely: a native Rust editor built for raw speed, fully open source, that treats AI as a composable layer rather than a built-in product.

If you're a solo developer or indie hacker deciding where to do your daily coding in 2026, these three tools represent genuinely different philosophies. Picking the wrong one costs you either money or productivity. Here's the honest breakdown.

Quick verdict

Editor	Price	Best for	Agent approach	Rating
Cursor	$20/month Pro	Developers who want the most mature AI IDE ecosystem	Built-in Agent mode, credit-based	⭐⭐⭐⭐⭐
Windsurf	$15/month Pro	Developers who want the most autonomous coding agent	Cascade (acts first, asks less)	⭐⭐⭐⭐⭐
Zed	Free / $20/month	Speed-obsessed developers and open source advocates	External agents via ACP, own API keys	⭐⭐⭐⭐

None of these is objectively better. The right choice depends on whether you want a polished all-in-one AI IDE (Cursor or Windsurf) or a blazing-fast editor where you compose your own AI workflow (Zed).

Cursor: the ecosystem standard

Cursor is a VS Code fork with AI deeply integrated into the editing experience. It's been around long enough that most edge cases have been solved and documented by its large community.

Price: Free (limited), Pro $20/month, Pro+ $60/month, Ultra $200/month, Business $40/seat/month.

The Pro plan gives you a $20 monthly credit pool for frontier models. Auto mode is effectively unlimited and uses cost-efficient models automatically. Credits only deplete when you manually select expensive frontier models like Claude Sonnet or GPT-5.4 on heavy tasks.

What sets it apart:

Agent mode (upgraded from Composer in 2026) handles multi-file, multi-step tasks. It pauses more often for confirmation than Windsurf's Cascade, which is either a feature or a bug depending on how much oversight you want. For developers who like reviewing changes before they're applied, this behavior is preferable.

The community is the real differentiator. Over a million paying users, thousands of Discord threads, and a growing ecosystem of tools, plugins, and workflows built on top of Cursor. When you hit an edge case at 2am (and you will), there's a good chance someone already solved it.

Full codebase context awareness, background agents that run while you work on something else, MCP server support, and bring-your-own-API-key mode are all included.

Who should use it: Developers who want the most production-tested AI IDE with the largest community. If you're working on complex multi-file codebases and want turnkey AI workflows without assembling anything yourself, Cursor is the right call.

The honest con: The credit system confused a lot of people when Cursor introduced it in mid-2025, and billing surprises still happen when developers manually reach for expensive frontier models. Turn on spend limits before you start. The VS Code dependency also means you're locked into that editor paradigm. If you've ever wanted something faster or lighter, Cursor won't scratch that itch.

Windsurf: the most autonomous agent

Windsurf is Codeium's flagship IDE, built on VS Code like Cursor but with a fundamentally different AI philosophy. Where Cursor's agent asks for confirmation regularly, Windsurf's Cascade acts first and asks less.

Price: Free (25 prompt credits/month, unlimited Tab autocomplete), Pro $15/month, Teams $30/user/month, Enterprise $60/user/month.

Note: Windsurf updated its pricing in March 2026 with a new quota-based system. The Pro tier was $15/month at launch and some sources report it moved to $20 after the pricing restructure. Verify current pricing at windsurf.com/pricing before subscribing.

Tab autocomplete is unlimited on every plan including free, which means autocomplete-heavy workflows cost nothing extra regardless of tier.

What sets it apart:

Cascade is Windsurf's core product decision. It's more aggressive by default: it executes multi-step tasks with less interruption than Cursor's agent. For developers who want to describe a task and come back to reviewed results rather than approving every step, Cascade feels more natural.

The SWE-1 and SWE-1.5 proprietary models consume zero credits, which means you can run Cascade extensively without eating into your credit allocation if you're comfortable with Windsurf's own models. Switching to Claude Sonnet or GPT-5.4 within Windsurf does consume credits.

Windsurf also runs natively in JetBrains IDEs via a plugin, which matters if you're a Laravel developer who prefers PhpStorm over VS Code.

Who should use it: Developers who want the most hands-off autonomous coding experience. If you want to describe a feature and have the agent implement it across multiple files with minimal interruption, Windsurf's Cascade delivers this more smoothly than Cursor's agent in most workflows.

The honest con: The proprietary SWE models are good but not at the same level as Claude Sonnet or GPT-5.4 for complex reasoning tasks. Using frontier models within Windsurf means managing credits, which reintroduces the same unpredictability as Cursor's system. The community is smaller than Cursor's, which means less community support for edge cases.

Zed: a completely different approach

Zed is not a VS Code fork. It was built from scratch in Rust by the former Atom team, and the architecture shows: it starts in 0.12 seconds, renders at 120fps, and uses under 300MB of memory even in large codebases. VS Code with a typical extension set exceeds 1GB.

Price: Free (50 hosted AI prompts/month, unlimited if you bring your own API keys), Pro $20/month (500 hosted prompts). Crucially: if you already pay Anthropic, OpenAI, or Google directly for API access, Zed's AI features cost you nothing extra. You bring your key, Zed uses it.

What sets it apart:

Zed treats AI as a composable layer rather than a bundled product. The Agent Client Protocol (ACP) lets you connect Claude Code, OpenAI's Codex, or other external agents directly into Zed. You're not locked into Zed's built-in agent. You can run Claude Code inside Zed while still benefiting from Zed's editing speed.

The open source angle is real. The core editor is GPL-licensed, GPUI (the GPU rendering framework) is Apache 2, and the Zeta2 edit prediction model is fully open weight. Security teams can audit every line. Privacy-conscious developers can keep code completely off third-party servers by using local models via Ollama.

Real-time multiplayer collaboration (voice + CRDT cursors) is built in. Git integration is native. The Vim mode is considered one of the best implementations available.

Who should use it: Developers who feel VS Code or Cursor is too slow, who want genuine open source, or who prefer composing their own AI workflow rather than buying a packaged one. Also the right choice if you already pay for Claude or Gemini API access and don't want to pay a second subscription for AI in your editor.

The honest con: Zed is macOS and Linux only with a stable release. Windows is in beta and not recommended for production workflows yet. The AI agent experience requires more assembly than Cursor or Windsurf. You get composability but not polish. The extension ecosystem is smaller than VS Code's, which is still the dominant gravitational force for most teams.

Head-to-head: what actually matters

Autonomy level: Windsurf's Cascade is the most autonomous out of the box. Cursor's agent pauses more for confirmation. Zed's built-in agent is capable but the real power is connecting external agents via ACP.

Price at $0: Zed wins. 50 hosted prompts plus unlimited if you bring your own API key. Cursor and Windsurf free tiers are for evaluation only.

Price at paid tier: Windsurf Pro at $15/month undercuts both. Cursor and Zed Pro are both $20/month.

Editor performance: Zed wins by a wide margin. 120fps rendering, 0.12 second launch time, sub-300MB memory. Cursor and Windsurf are Electron-based VS Code forks with the corresponding performance characteristics.

Community and ecosystem: Cursor wins. The largest user base, most Discord activity, and most community-built tools.

Open source: Zed wins. Core editor is GPL, fully auditable, local model support via Ollama.

Platform support: Cursor and Windsurf win. Both run on macOS, Linux, and Windows. Zed's Windows support is still in beta.

MCP support: All three support MCP servers. Zed's ACP standard is an open protocol that lets any agent integrate without bespoke connectors.

How to choose

You want the most mature, community-backed AI IDE: Cursor. The ecosystem, community resources, and production track record are unmatched. Worth the $20/month if you work in a VS Code environment already.

You want the most autonomous agent that acts without interruption: Windsurf. Cascade's default behavior matches how most developers actually want an agent to work. The $15/month price makes it easier to justify than Cursor.

You already pay for Claude or Gemini API access: Zed. You get a world-class editor and use your existing AI subscription inside it. Net cost is $0 on top of what you're already paying.

You want open source and privacy first: Zed. Run local models via Ollama, keep code off third-party servers, audit every line of the editor code.

You're on Windows: Cursor or Windsurf. Zed's Windows support is not production-ready yet.

You want the fastest editor and don't mind assembling your AI workflow: Zed. Nothing else comes close on raw editor performance.

FAQ

Can I use Cursor, Windsurf, and Zed together?

Yes. Many developers use Zed as their primary editor for its speed, and run Claude Code or Cursor for specific heavy autonomous tasks. They're not mutually exclusive. Zed's ACP specifically makes it designed to work alongside other agents.

Does Windsurf's Cascade work with Claude and GPT models?

Yes. Windsurf supports Claude Sonnet, GPT-5.4, Gemini, and its own SWE models. Using frontier models consumes credits from your monthly allocation. Windsurf's own SWE models consume zero credits.

Is Zed's free tier actually usable for daily work?

With your own API key, yes. Fully usable for daily work. The 50 hosted prompt limit applies only if you use Zed's AI quota. Bring your own Anthropic or OpenAI key and there are no Zed-imposed limits on AI usage. You just pay your API provider directly.

How does Cursor's credit system work in practice?

Auto mode is unlimited and uses cost-efficient models. The $20 credit pool depletes only when you manually select expensive frontier models. Most developers on the $20 Pro plan never exhaust their credits in normal usage. Problems happen when developers switch to Claude Opus or similar on every request.

Is Windsurf's JetBrains plugin as good as the standalone IDE?

It brings Cascade into JetBrains, which is meaningful for PhpStorm and IntelliJ users. It's not identical to the full Windsurf IDE experience but covers the core agent functionality. Worth trying if you're a Laravel developer who prefers JetBrains tooling.

Bottom line

If you're choosing one tool today, the decision mostly comes down to what you value:

Community and ecosystem: Cursor. Autonomous agent that acts without handholding: Windsurf. Speed, open source, and composability: Zed.

All three are genuinely good tools in 2026. The AI code editor space has matured enough that any of them will make you more productive than working without one. The question is which philosophy matches your workflow.

Comparing AI coding tools? Also read: Cursor vs GitHub Copilot vs Claude Code in 2026 and ChatGPT Pro $100 vs Claude Max vs Cursor: Which Subscription Wins?.

ShipFast vs Larafast vs SaaSykit for Indie Hackers in 2026: Which SaaS Boilerplate Is Worth It?

DevToolsPicks — Sun, 12 Apr 2026 05:51:37 +0000

Every SaaS boilerplate promises the same thing: stop wiring up auth, payments, and email from scratch and start building your actual product. The pitch is real. Auth alone takes 20+ hours if you're doing it properly. Add Stripe webhooks, email templates, an admin panel, and a landing page and you're looking at 80-100 hours before you've written a single line of business logic.

The question isn't whether a boilerplate is worth it. For most solo developers, it clearly is. The question is which one fits your stack, your project, and your budget.

ShipFast at $199 is the NextJS option. Larafast at $169 and SaaSykit at $179 are the Laravel options. Here is what actually matters when choosing between them.

Quick verdict

Boilerplate	Price	Stack	Best for	Rating
ShipFast	$199 one-time	NextJS, React, Tailwind	NextJS developers who want the fastest launch	⭐⭐⭐⭐⭐
Larafast	$169–$199 one-time	Laravel, TALL or VILT	Laravel developers who want stack flexibility	⭐⭐⭐⭐⭐
SaaSykit	$179 one-time	Laravel, TALL (Livewire)	Laravel developers who want built-in SaaS metrics	⭐⭐⭐⭐

The first decision is not ShipFast vs Larafast. It's NextJS vs Laravel. Pick the boilerplate that matches the framework you already know. Switching frameworks to use a boilerplate is never worth it.

If you're a NextJS developer: ShipFast is the default choice in its category.
If you're a Laravel developer: Larafast and SaaSykit are both strong, with different trade-offs.

ShipFast: the NextJS standard

ShipFast was built by Marc Lou, one of the most prolific indie hackers in the space. He's shipped 27+ products using it, which means it's been tested in production across dozens of different use cases rather than built as a demo project.

Price: $199 one-time, unlimited projects.

What you get:

NextJS App Router (JavaScript and TypeScript options)
Auth: Google OAuth, magic link email login, protected API routes
Payments: Stripe (subscriptions and one-time) + Lemon Squeezy
Database: MongoDB or Supabase (PostgreSQL)
Email: Mailgun and Resend
UI: Tailwind CSS components, pre-built landing page, SEO blog
Community: 8,000+ active users, 5,000+ in Discord

The community is genuinely the differentiating factor. At 8,000+ users, most integration edge cases have already been solved and documented in the Discord. When you hit a problem combining Supabase auth with Stripe webhooks at 2am before your launch, there is probably a thread about it.

The documentation is detailed. Setup to deployed SaaS with real auth and billing reportedly takes under a day for most developers.

Who should use it: NextJS developers who want to ship fast and don't want to figure out boilerplate architecture decisions from scratch. Marc Lou uses this for every product he ships, which is the strongest possible evidence that it works in real production scenarios.

The honest con: ShipFast is deliberately minimal on features beyond the boilerplate essentials. There's no multi-tenancy, no built-in team management, no SaaS metric tracking. If your product needs team workspaces or complex subscription analytics, you'll be building those yourself. ShipFast is also a NextJS product. Deploying to Vercel is the natural path, which adds ongoing hosting costs at scale that you don't have with a self-hosted Laravel VPS.

Larafast: the Laravel indie hacker pick

Larafast occupies the same market position as ShipFast but for Laravel developers. If you're a PHP developer and the TALL or VILT stack is your natural environment, this is the equivalent starting point.

Price: $169 one-time (TALL stack) or $199 (VILT stack). One purchase, unlimited projects.

What you get:

Laravel with TALL stack (Tailwind, Alpine.js, Livewire, Laravel) or VILT (Vue, Inertia.js, Laravel, Tailwind)
FilamentPHP admin panel (user management, subscription management, blog, settings)
Payments: Stripe, Lemon Squeezy, Paddle
Auth: email/password, Google OAuth, social login
SEO tools and blog system built in
Landing page components ready to go

The dual-stack choice is Larafast's clearest advantage over every other Laravel boilerplate. If you prefer Vue with Inertia.js you get that. If you prefer Livewire you get that. No other Laravel starter kit gives you the option at purchase. This matters because switching stacks mid-project is expensive, and choosing between Livewire and Vue/Inertia is a real decision point for Laravel developers.

FilamentPHP is the dominant admin framework in the Laravel ecosystem in 2026. Getting it pre-configured with user management, subscription controls, and a blog resource saves several days of work.

Three payment providers out of the box is also worth calling out. Stripe for US/EU founders, Lemon Squeezy for MoR simplicity, Paddle for digital goods. You can pick based on your situation without writing integration code.

Who should use it: Laravel developers building SaaS products who want to launch quickly without rebuilding the same auth/billing/admin setup for every project. Particularly valuable if you ship multiple products per year and want one solid foundation to reuse.

The honest con: Larafast is maintained by a solo developer, which means feature additions and bug fixes depend on one person's availability. The community is smaller than ShipFast's (340+ developers compared to 8,000+). For straightforward use cases this doesn't matter much. For edge cases and complex integrations, you're more likely to be on your own. Documentation quality is solid but not as comprehensive as ShipFast's.

SaaSykit: best billing intelligence for Laravel

SaaSykit is the other serious Laravel contender. It's built on the TALL stack with FilamentPHP like Larafast, but with one feature that stands out: built-in SaaS business metrics.

Price: $179 one-time, unlimited projects.

What you get:

Laravel + TALL stack (Livewire, AlpineJS, Tailwind)
FilamentPHP admin panel
Payments: Stripe, Paddle, Lemon Squeezy, and Creem
Built-in MRR, churn rate, ARPU, and trial conversion dashboards
Multi-tenancy version available separately
Fully translatable (i18n out of the box)
Comprehensive automated test coverage

The SaaS metrics dashboard is the reason SaaSykit exists as a separate product from Larafast. Out of the box, you can see your MRR, churn, ARPU, new vs churned revenue, and trial conversion rates inside the FilamentPHP admin. Building something equivalent from scratch takes time and you'd typically pay $49/month to a tool like Baremetrics for the same visibility. SaaSykit builds it into your own codebase.

Creem support is also notable SaaSykit is the only boilerplate on this list that supports Creem as a payment processor alongside Stripe, Paddle, and Lemon Squeezy.

The test coverage is another genuine differentiator. Most SaaS boilerplates ship with minimal testing. SaaSykit documents its test coverage across critical components, which matters more as your product grows and you start refactoring core features.

Who should use it: Laravel developers who are building a product they plan to run seriously for multiple years, where understanding your subscription metrics from day one matters. Also the right choice if you need multi-tenancy (team workspaces with seat-based billing), which SaaSykit Tenancy handles.

The honest con: No VILT option. You get TALL only. If your preferred Laravel frontend is Vue with Inertia.js, Larafast is the better fit. The community is smaller than both ShipFast and Larafast in terms of public presence. The $179 price puts it between the two and the feature set justifies it, but only if you'll actually use the metrics dashboard.

Head-to-head: what actually matters

Framework lock-in: This is the real decision. ShipFast is NextJS. Larafast and SaaSykit are Laravel. Choose based on your stack, not on feature lists.

Payment provider coverage: SaaSykit wins with Stripe, Paddle, Lemon Squeezy, and Creem. Larafast covers Stripe, Lemon Squeezy, and Paddle. ShipFast covers Stripe and Lemon Squeezy.

Admin panel: Both Laravel options use FilamentPHP, which is the right choice in 2026. ShipFast has no dedicated admin panel you'd build or integrate one yourself.

Community: ShipFast wins decisively at 8,000+ users. This translates to faster answers to edge case questions and a larger ecosystem of plugins and extensions built on top of it.

Deployment cost at scale: ShipFast on Vercel means your hosting bill grows with traffic. Laravel on a VPS (your own Hetzner or DigitalOcean server with Laravel Forge or Ploi) gives you more predictable costs and more control as your product grows.

Built-in SaaS metrics: SaaSykit is the only option with MRR/churn dashboards out of the box. For everyone else, you're adding this later.

Multi-tenancy: SaaSykit Tenancy is the only one with a dedicated multi-tenant version. ShipFast and Larafast don't include this.

How to choose

You're a NextJS developer: ShipFast. There's no serious Laravel-equivalent in the NextJS space with the same community size and production track record. The $199 is worth it for the Discord alone.

You're a Laravel developer who prefers Vue/Inertia.js: Larafast VILT at $199. The stack flexibility is the key feature and nothing else on this list offers it for Laravel.

You're a Laravel developer who prefers Livewire and wants SaaS metrics built in: SaaSykit at $179. The billing intelligence dashboard pays for itself the first time you're trying to diagnose why your churn is spiking.

You're a Laravel developer, Livewire is fine, and you're shipping multiple small products: Larafast TALL at $169. The lower price and solid foundation make it the right call for high-volume product shipping.

You need multi-tenant SaaS (team workspaces, seat billing): SaaSykit Tenancy. Nothing else here handles this out of the box.

FAQ

Is a SaaS boilerplate actually worth $169-$199?

If you're billing your time at any professional rate, yes. The setup work a boilerplate replaces auth, payments, email, admin panel, landing page takes 60-100 hours to build properly from scratch. Even at $25/hour, that's $1,500-$2,500 in time saved. The boilerplate cost is irrelevant at that scale. The real question is whether the boilerplate fits your stack and project requirements.

Can I use these boilerplates for client work?

All three are one-time purchases with unlimited project licenses. You can build multiple projects, including client work, without additional licensing fees. Verify the specific license terms on each product's site before commercial client deployments.

ShipFast vs just using a NextJS starter template from the official docs?

The official NextJS docs provide auth and basic Stripe setup, but you're assembling it yourself. ShipFast gives you auth, payments, email, a landing page, a blog, and UI components pre-integrated and tested across 27+ real products. The free option works if you have time. ShipFast works if you don't.

What if Larafast stops being maintained?

This is a real concern with any solo-maintained boilerplate. Larafast gives you the full codebase you're not locked into a service, just a starting point. If maintenance stops, your existing projects keep running and you own all the code. The risk is that future Laravel version compatibility updates may require your own work rather than a simple update pull.

Do I need to know Laravel well to use Larafast or SaaSykit?

Yes. Both are for developers who already work in Laravel. They're not for learning Laravel they're for skipping the setup work that Laravel developers do repeatedly across projects. If you're new to Laravel, spend time with the framework first before buying a boilerplate.

Bottom line

The boilerplate decision is mostly a stack decision. NextJS developers should look at ShipFast first. Laravel developers should decide between Larafast and SaaSykit based on whether the VILT stack option or the built-in metrics dashboard matters more to their next project.

All three are legitimate time-savers for solo developers who ship real products. None of them replace the need to build your actual product they just eliminate the month of setup work before you can start.

Building your Laravel SaaS stack? Also read: Laravel Forge vs Ploi vs Coolify: Which Should Solo Devs Use? and Lemon Squeezy vs Stripe vs Paddle for Solo Devs.

Best Typeform Alternatives for Indie Hackers in 2026 (Honest Picks)

DevToolsPicks — Sat, 11 Apr 2026 05:11:50 +0000

Typeform's free plan gives you 10 responses per month. Ten. If you build a waitlist form and share it in one Reddit thread, you'll hit that limit before lunch.

The Basic plan at $29/month buys you 100 responses. That's $0.29 per response at full capacity, before you've even launched. For a solo developer or indie hacker validating an idea, that pricing makes no sense.

The beautiful conversational UX is real. But beautiful doesn't pay the AWS bill. Here are five alternatives that give you the form features you actually need without the response-limit anxiety.

Quick verdict

Tool	Price	Best for	Responses (free)	Rating
Tally	Free / $29/month	Solo builders who want everything for free	Unlimited	⭐⭐⭐⭐⭐
Fillout	Free / $19/month	Teams needing Notion/Airtable sync	1,000/month	⭐⭐⭐⭐⭐
Formbricks	Free / $49/month	Devs who want open source and self-hosting	1,000/month	⭐⭐⭐⭐
Jotform	Free / $34/month	Feature-heavy forms with 300+ templates	100/month	⭐⭐⭐
Google Forms	Free	Internal use, quick surveys	Unlimited	⭐⭐⭐

The short version: Tally is the default pick for most indie hackers. Fillout wins if you live in Notion or Airtable. Formbricks is the move if you care about data ownership or need in-app surveys.

Why people leave Typeform

Before getting into alternatives, it helps to understand what actually frustrates people about Typeform, because not every alternative solves every problem.

The 10-response free plan is the most obvious issue. Almost no other form builder gates free usage this aggressively. Tally, Fillout, and Formbricks all offer at least 1,000 free responses. Google Forms is unlimited.

Less discussed: Typeform counts partial submissions against your response limit. Someone who opens your form and fills in one field then abandons it still eats into your quota. On a $29/month plan with 100 responses, a 50% completion rate means you're paying $0.58 per useful lead.

And CAPTCHA (basic bot protection) is not available on any Typeform Core plan. You have to upgrade to Growth Essentials at $199/month to get it. Run any paid ads to a Typeform and you'll be paying for bot submissions that you can't filter out.

Tally: the default pick for indie hackers

Tally was built specifically to fix the Typeform pricing problem. The founders were frustrated with expensive form builders and built something genuinely unlimited on the free plan.

What it costs: Free forever (with fair use guidelines). Pro at $29/month. Business at $89/month.

What the free plan actually includes: Unlimited forms, unlimited submissions, conditional logic, file uploads, Stripe payments, e-signatures, and Notion/Google Sheets integration. That's not a trial. That's the free plan.

The interface is Notion-like. You type and forms build around you. It clicks immediately if you already use Notion daily. Building a multi-step lead capture form with conditional logic takes maybe 10 minutes the first time.

Pro at $29/month adds: Custom domains, Tally branding removal, custom CSS, partial submission capture, extended analytics, and team collaboration. For most indie hackers, the free plan covers everything until you're generating real revenue.

Who should use it: Solo developers and indie hackers who need forms for waitlists, feedback collection, payment flows, or lead generation. If you're validating an idea and need to collect 500 emails this week, Tally is the obvious choice.

The honest con: Conditional logic is less visual than Typeform. If you're building complex multi-path surveys with many branches, Tally requires more setup effort. The referral program exists but commission terms aren't publicly disclosed, so there's no affiliate angle here for anyone evaluating it as a business.

Fillout: best for Notion and Airtable users

Fillout sits in the middle of the market. More powerful than Tally on integrations, more affordable than Typeform at every tier.

What it costs: Free plan: unlimited forms, unlimited seats, 1,000 responses/month. Starter at $19/month (2,000 responses). Pro at $49/month (5,000 responses, custom branding). Business at $89/month (unlimited responses).

The standout feature: Bidirectional sync with Airtable and Notion. Most form builders write to a spreadsheet. Fillout reads from and writes to your Airtable base or Notion database in real time. If your project already lives in Notion, Fillout forms become an extension of that workspace rather than a separate data silo.

The free tier is genuinely useful at 1,000 responses/month. That covers most indie hackers for the first few months of a real launch without paying anything.

Who should use it: Builders whose workflow is centered on Notion or Airtable. Also worth considering if you need a step up from Google Forms in design quality without jumping to Typeform prices.

The honest con: Branding removal starts at Pro ($49/month), which is more expensive than Tally's $29/month for the same feature. If your primary need is removing "Powered by Fillout" from customer-facing forms, Tally is cheaper for that specific outcome.

Formbricks: for developers who want data control

Formbricks is the only open-source option on this list and the only one you can genuinely self-host. If you've ever worried about form responses living on someone else's servers, this solves it.

What it costs: Free cloud plan: unlimited surveys, 1,000 responses/month, API access, conditional logic, in-app surveys. Startup at $49/month: 5,000 responses, no branding, attribute-based targeting. Self-hosted: free forever, no response limits.

The standout feature: In-app surveys. Typeform, Tally, and Fillout are all link-based forms. You share a URL and users fill it out. Formbricks can trigger surveys inside your actual product based on user actions, page visits, or custom events. For a SaaS product, this means showing a feedback prompt after a user completes a specific workflow, not just hoping they click a link.

The self-hosting option is real and genuinely supported. You run it on your own VPS with Docker, data stays on your infrastructure, and there are no per-response fees. For indie hackers already running a VPS (which you should be if you're a developer), the incremental hosting cost is close to zero.

Who should use it: Developers building SaaS products who need in-app feedback, or anyone working in regulated industries (healthcare, fintech) where data sovereignty matters. Also the right pick if you want zero vendor lock-in.

The honest con: More setup effort than any other tool on this list. The self-hosted path requires Docker knowledge and server management. The cloud free plan's 1,000 response limit is identical to Fillout but the paid Startup tier at $49/month is more expensive than Tally or Fillout for similar features.

Jotform: the feature-complete option

Jotform has been around since 2006. That longevity shows: 300+ templates, 100+ integrations, payment collection via Stripe and PayPal, approval workflows, and a drag-and-drop builder that works for people who are not technical.

What it costs: Free (Starter): 5 forms, 100 submissions/month. Bronze at $34/month annual: 25 forms, 1,000 submissions. Silver at $39/month annual: 50 forms, 2,500 submissions. Gold at $99/month annual: 100 forms, 10,000 submissions.

The case for Jotform: If you need something built and working in 10 minutes without touching any logic or settings, Jotform's template library is unmatched. Medical intake forms, event registration, payment collection, conditional multi-page surveys. It's all there, ready to go.

Who should use it: Non-technical founders or indie hackers who need forms that go beyond simple lead capture. Things like order forms, booking flows, or data collection for client work. The template library saves real time.

The honest con: The free plan's 5-form limit is genuinely restrictive. And every plan below Enterprise is single-user only. If you want a team member to also manage forms, you're either paying for two accounts or jumping to custom Enterprise pricing. Tally and Fillout both allow team collaboration at much lower price points. The per-form and per-submission limits across all tiers make budgeting unpredictable as usage scales.

Google Forms: when free is the only budget

Google Forms deserves a mention because it's completely free, unlimited, and already in your Google account.

Unlimited forms, unlimited responses, basic conditional logic, Google Sheets auto-sync, and zero setup. For internal tools, quick customer surveys, or collecting beta signups from people you're already in contact with, it does the job.

The honest con: You cannot remove Google branding, styling options are minimal, there's no payment collection, and the visual design signals "I didn't spend money on this" to anyone who sees it. For customer-facing forms at a serious product, it affects trust. Use it for internal use cases and upgrade to Tally when you need something that looks professional.

How to choose

Starting from zero, need forms now: Tally free plan. No setup friction, unlimited responses, looks clean.

Your data lives in Notion or Airtable: Fillout. The bidirectional sync is a genuine workflow advantage that Tally doesn't match.

Building a SaaS product and need feedback inside the product: Formbricks. Nothing else on this list does in-app surveys with targeting.

Non-technical, need complex forms fast, willing to pay: Jotform. The template library justifies the price for the right use case.

Internal tools only, budget is zero: Google Forms. Don't overthink it.

What you don't need to do: Pay $29/month to Typeform for 100 responses when you're still validating whether anyone wants your product.

FAQ

Is Tally really free with unlimited responses?

Yes, within fair use guidelines. Tally is a bootstrapped EU company and the fair use policy exists to prevent abuse at very high volumes, but for most indie hackers building real products, the free plan covers everything. Their Pro plan at $29/month is still cheaper than Typeform's Basic plan and gives you unlimited responses instead of 100.

Can I use Tally or Fillout for payment collection?

Both support Stripe payments on the free plan. Tally includes Stripe on free; Fillout includes payments on free too. Typeform only supports payments on Business tier ($99/month). This alone makes Tally and Fillout significantly better for early-stage products collecting pre-orders or one-time payments.

Is Formbricks difficult to self-host?

If you're comfortable with Docker and already run a VPS, the setup takes about 30 minutes following their documentation. If you've never set up a Docker container, use the cloud free plan instead and self-host later when you need the control.

What's wrong with Google Forms?

Nothing, for the right use case. The limitations are: no payment collection, no custom branding removal, basic styling only, and no conditional logic beyond simple branching. For internal operations or early beta feedback from friendly users, it works well. For customer-facing forms at a product you want people to take seriously, upgrade to Tally.

Does Typeform's free plan actually work for anything?

10 responses per month is enough to test your form before sharing it. That's about it. If you share a form publicly on any channel, you'll exhaust the free plan in hours.

Bottom line

Typeform built something genuinely beautiful. But beautiful forms at $29/month for 100 responses are a poor match for indie hackers who are still figuring out whether their product has a market.

Tally gives you everything you need for free and then charges a reasonable $29/month when you actually need custom domains and branding removal. That's the right order of operations for early-stage products.

Start with Tally. Add Fillout if your workflow is Notion-first. Consider Formbricks when data ownership becomes a real requirement. Revisit Typeform if and when conversion rates justify the cost.

Building your SaaS stack? Also read: Best Zapier Alternatives for Solo Developers in 2026 and Polar vs Lemon Squeezy vs Creem in 2026.

ChatGPT Pro $100 vs Claude Max vs Cursor: Which AI Coding Subscription Is Worth It for Indie Hackers in 2026?

DevToolsPicks — Fri, 10 Apr 2026 05:36:26 +0000

OpenAI launched a $100/month ChatGPT Pro tier yesterday. It is not subtle about who they are going after. TechCrunch quoted an OpenAI spokesperson directly: the new tier "delivers more coding capacity per dollar across paid tiers" compared to Claude Code. That is a direct shot at Anthropic.

So now you have two $100/month AI subscriptions fighting for the same developer wallet. Claude Max 5x has been at $100 since late 2024. OpenAI just matched it, priced identically, aimed at the same audience. Meanwhile Cursor Pro sits at $20 and does something different enough to deserve its own spot in this comparison.

If you are a solo developer or indie hacker trying to figure out where your subscription budget goes this month, here is the honest breakdown.

Quick verdict

Tool	Price	Best for	Coding agent	Rating
ChatGPT Pro $100	$100/month	Heavy Codex users switching from Claude	Codex (cloud + local)	⭐⭐⭐⭐
Claude Max 5x	$100/month	Developers already in the Claude ecosystem	Claude Code	⭐⭐⭐⭐⭐
Cursor Pro	$20/month	Daily IDE users who want AI inside their editor	Agent mode (in-editor)	⭐⭐⭐⭐

If you are starting from scratch and want the most capable standalone AI coding agent, Claude Max still wins. If you are already deep in the OpenAI ecosystem or have been hitting Claude's usage limits, the new ChatGPT Pro tier is worth a real look. Cursor is a different product entirely and the right choice if you want AI inside your editor rather than running tasks in a terminal.

ChatGPT Pro $100: what you actually get

OpenAI's full plan lineup now looks like this: Free (with ads), Go ($8/month, with ads), Plus ($20/month), Pro $100 (new), Pro $200 (still exists, quietly).

The $100 tier gives you:

5x more Codex usage than the $20 Plus plan
All Pro features including GPT-5.4 Pro model access
Unlimited access to Instant and Thinking models
Codex running both locally on your machine and as cloud tasks
Bonus through May 31: 10x Codex usage instead of 5x during the launch promotion

Codex is OpenAI's agentic coding tool. It handles multi-file edits, runs shell commands, writes and executes tests, and works on tasks in the background while you do other things. Think of it as OpenAI's answer to Claude Code.

The naming is a mess. There are now two ChatGPT Pro plans, both called "Pro." The $100 one is for 5x usage, the $200 one is for 20x. OpenAI published an explainer about their own naming, which tells you everything about how clearly this was thought through.

The real reason this exists: Anthropic's Claude Code has been dominant in the agentic coding space. Anthropic's ARR reportedly topped $30 billion, driven largely by Claude Code adoption. OpenAI needed a competitive $100 tier. The timing is not a coincidence.

Who should use it: Developers who have been using ChatGPT Plus and hitting Codex limits regularly. Also worth considering if you were using Claude Code through a third-party harness that Anthropic restricted in early April and you are looking to switch.

The honest con: Codex is good but Claude Code has had a meaningful head start in the agentic coding market. The ecosystem, documentation, and community of tips around Claude Code is more mature. OpenAI's 10x launch promo through May 31 is designed to get you hooked before the limits drop back to 5x. Be aware of that before you cancel anything.

Claude Max 5x: the incumbent

Anthropic launched the Max plan in late 2024. At $100/month, Max 5x gives you:

5x the usage of Claude Pro ($20/month)
Claude Code in the terminal, web, and desktop
Full Opus 4.6 access (1 million token context window, agent teams)
Persistent memory across conversations
Priority access during peak hours
Early access to new features

Claude Code is Anthropic's terminal-native coding agent. It reads your full project, plans changes across files, executes shell commands, handles multi-file refactors, and can run autonomous sessions. If you have used it for more than a few hours on a real project, you know why 3 million people a week use Codex and more are on Claude Code. They are genuinely different from autocomplete tools.

The Max 20x plan at $200/month takes limits off the table entirely for most developers. At that level, full-day coding sessions, parallel agent tasks, and heavy Opus 4.6 usage stop being a concern.

One developer tracked their usage over 8 months and found the Max plan saved over 93% compared to equivalent API costs. At heavy usage levels, the math strongly favors the subscription.

Who should use it: Developers who use AI coding tools as their primary development workflow and hit Pro rate limits more than twice a week. If you are building SaaS products solo and Claude Code is your main tool, Max 5x is the right call.

The honest con: Anthropic made a controversial move in early April blocking Claude subscriptions from powering third-party agentic harnesses. If you were relying on something like OpenClaw to extend Claude's functionality, that access is now restricted. This is part of why OpenAI's $100 tier is landing on fertile ground right now.

Cursor Pro: a different question entirely

Cursor is not a standalone coding agent. It is an AI-powered IDE built on VS Code. That distinction matters.

Cursor Pro at $20/month gives you:

$20 monthly credit pool for frontier model access (Claude Sonnet, GPT-5.4, Gemini)
Agent mode for multi-file edits inside the editor
Auto mode (effectively unlimited, uses lower-cost models automatically)
Background agents that run tasks while you work on something else
Full codebase context awareness

The credit system confused a lot of people when Cursor introduced it in mid-2025. The short version: if you stick to Auto mode, Pro is effectively unlimited for most use cases. Credits only deplete when you manually select expensive frontier models on heavy tasks. Most developers never exhaust their monthly pool.

Cursor Pro+ at $60/month triples your credit pool. Ultra at $200/month gives you $400 in credits.

Who should use it: Developers who want AI integrated directly into their editor experience. Cursor is better than Claude Code if you want to stay in an IDE, use multiple AI models interchangeably, and prefer in-context editing over terminal-based agents. It is also the right answer if $20/month is your actual budget.

The honest con: Cursor requires you to abandon your current editor. If you are on JetBrains, real Vim, or anything that is not VS Code-based, the switch cost is real. The credit system also caused genuine billing surprises when it launched. Turn on spend limits before you start using frontier models manually.

Head-to-head: the dimensions that actually matter

Coding agent capability: Claude Code and Codex are both serious agentic tools. Claude Code has been in the market longer, has more community resources, and handles complex multi-file refactors well. Codex is newer as a standalone product but OpenAI has invested heavily. Cursor's agent mode is capable but lives inside an editor rather than running fully autonomously.

Usage limits at $100: Both Claude Max 5x and ChatGPT Pro $100 offer 5x their base plan. The difference is what the base plan actually allows. OpenAI is "rebalancing" Plus limits as part of this launch, which effectively means Plus users get slightly less so the $100 tier looks more valuable. Through May 31, ChatGPT Pro $100 users get 10x, but that is temporary.

Model access: Claude Max 5x includes Opus 4.6, which is Anthropic's best model. ChatGPT Pro $100 includes GPT-5.4 Pro. Both are flagship-tier models at this price point.

Price: ChatGPT Pro $100 and Claude Max 5x are identical at $100/month. Cursor Pro is $20. The question is whether you need a standalone coding agent ($100 tier tools) or IDE-integrated AI ($20 with Cursor).

The competitive context: OpenAI explicitly launched this tier to compete with Anthropic. One person emailed Tibo at Codex saying they would cancel their Claude Max plan the moment a $100 Codex option launched. They did. That is the audience OpenAI is targeting. If you are loyal to neither ecosystem, the honest advice is to trial both and see which agent works better on your actual codebase.

Which one should you use?

Use ChatGPT Pro $100 if: You are currently on ChatGPT Plus and hitting Codex limits, you have recently switched away from Claude due to the third-party harness restrictions, or you want to try Codex during the 10x promo period (through May 31) before committing.

Use Claude Max 5x if: Claude Code is already your primary development tool and you hit Pro rate limits regularly. The ecosystem is more mature, the community is larger, and Opus 4.6 is genuinely excellent at complex multi-file work. Do not switch because of one frustrating rate limit day.

Use Cursor Pro if: You want AI deeply integrated into your editor rather than running in a terminal. $20/month is a real number for your budget. You prefer switching between Claude, GPT, and Gemini models depending on the task. Or you primarily need autocomplete and in-editor chat rather than long autonomous coding sessions.

Do not subscribe to two $100 tools. The temptation to stack Claude Max and ChatGPT Pro is real, especially during the Codex promo. That is $200/month on two tools that largely overlap. Pick one agent, stick with it for a month, and see where you hit friction.

FAQ

Is ChatGPT Pro $100 the same as the old Pro $200?

No. There are now two ChatGPT Pro plans. The new $100 tier gives you 5x Codex usage vs Plus. The $200 tier still exists and gives you 20x. OpenAI buried the $200 plan in their pricing page but confirmed it still works.

Does Claude Max include Claude Code?

Yes. Claude Code is included in Claude Pro ($20), Claude Max 5x ($100), and Claude Max 20x ($200). The difference between tiers is how much you can use it before hitting rate limits.

Is Cursor a replacement for Claude Code or Codex?

Not really. Cursor is an IDE with AI built in. Claude Code and Codex are standalone agents that work outside your editor. Many developers use Cursor for in-editor work and Claude Code for autonomous background tasks. They are complementary more than competitive.

Which has better limits, ChatGPT Pro $100 or Claude Max 5x?

Both offer 5x their base plan. The actual token counts vary by model and task type. Through May 31, ChatGPT Pro $100 temporarily offers 10x, but that promotional limit goes away.

Can I use my own API key with Cursor to avoid credit limits?

Yes. Cursor supports bring-your-own-API-key for OpenAI, Anthropic, and other providers. This bypasses Cursor's credit system entirely and you pay the model provider directly. Worth doing if you already have API access through your company.

Bottom line

OpenAI's $100 tier is a real product aimed at a real problem. The timing is strategic and the competition is good for developers. Both companies are now accountable to a clear price point.

If you are happy with Claude Code and not hitting limits constantly, stay on Max. Nothing in the Codex launch changes the quality of what you are already using. If you have been frustrated with Claude's recent restrictions or want to try Codex during the generous promo period, the $100 ChatGPT Pro tier is worth a month's trial.

For most solo developers on a realistic budget, Cursor Pro at $20 and Claude Pro at $20 together still costs less than one $100 tier subscription and covers most coding needs. The $100 tier is for developers who have outgrown that setup and need more.

Comparing AI coding tools? Also read: Cursor vs GitHub Copilot vs Claude Code in 2026 and Cursor 3's AI Agents Window: What Changed.

DEV Community: DevToolsPicks

How to Use Claude Code as a Solo Developer in 2026

Start with CLAUDE.md. Nothing else matters as much.

The three modes that matter

Slash commands worth building yourself

Which model to pick, and when

Pro vs Max vs API: the honest call

Four mistakes that cost real time

When NOT to reach for Claude Code

FAQ

Further reading

Final take

Codex Just Became a Full AI Agent: Is It Worth Switching From Claude Code?

Quick Verdict

What Actually Changed in Codex Today

Computer Use: The One Feature Claude Code Does Not Have

Codex vs Claude Code: Head to Head

Developer experience

Model quality

Pricing

Ecosystem and integrations

Regional availability

Who Should Actually Switch

Switch to Codex if:

Stay on Claude Code if:

Test both if:

Honest Cons of the New Codex

Pricing Table

FAQ

Is Codex now better than Claude Code overall?

Can I use Codex Computer Use in the EU?

Does Codex have MCP support?

Which is cheaper for heavy users?

Should I cancel my Claude Code subscription and switch to Codex?

What about Cursor 3?

What about Claude Code's new desktop app?

Final Recommendation

Best Sentry Alternatives for Indie Hackers in 2026 (Honest Picks)

Quick Verdict

The Problem With Sentry at Scale

GlitchTip: The Closest Drop-In Replacement

Bugsink: Single-Container Self-Hosting

Honeybadger: The Bundled Value Pick

Rollbar: The Cheapest Managed Paid Tier

Bugsnag: The Mobile-First Pick

How to Choose

FAQ

Does GlitchTip really work as a drop-in Sentry replacement?

What happens to my Sentry historical data if I migrate?

Is self-hosting Sentry itself a viable alternative?

Will my team notice the switch from Sentry?

What about PostHog for error tracking?

I still need session replay. What should I do?

Final Recommendation

Claude Opus 4.7 Just Launched: What Changed and Is It Actually Worth It?

Quick Verdict

What Actually Changed in Opus 4.7

Coding benchmarks

Reasoning and other benchmarks

Vision is the sleeper upgrade

New features in the API

New features in Claude Code

The Tokenizer Change Nobody Is Mentioning

Opus 4.7 vs GPT-5.4

Opus 4.7 vs Gemini 3.1 Pro

Who Should Switch to Opus 4.7 Right Now

Switch today:

Test before switching:

Do not bother yet:

Honest Cons

Pricing and How to Access It

Claude.ai (interactive)

API (claude-opus-4-7)

Cloud platforms

FAQ

Is Opus 4.7 faster than Opus 4.6?

Should I upgrade if I am already paying for Claude Max?

Is Opus 4.7 better than GPT-5.4 overall?

What is Claude Mythos and why is it in the benchmark chart?

Will my existing Opus 4.6 API code work with 4.7?

API (`claude-opus-4-7`)