All GitHub Copilot plans are now on usage-based billing

All GitHub Copilot plans are now on usage-based billing

Hello GitHub Community,

Usage-based billing is now in effect for all GitHub Copilot plans. Please refer to our initial announcement and today’s changelog for more details, and you’ll find an FAQ below.

Frequently Asked Questions

Can I buy as much additional usage as I want?

After you consume your included AI Credits, you can enable additional spend by updating your budget to keep using Copilot. However, we may limit how much additional usage users on individual plans can consume based on their usage patterns, billing history, and verification states for their GitHub account. When you hit the limit, Copilot features that require additional AI credits will pause. You can either wait for the next billing cycle or upgrade to a higher plan tier to continue using Copilot. We recommend upgrading to the next plan for more included usage and higher additional usage limits.

What are the additional usage limits? How much additional usage can users purchase?

The limits increase by plan, so the limits for users on a Pro+ plan will generally be higher than those for users on a Pro plan. They also vary depending on a user’s usage patterns, billing history, and verification states for their GitHub account. These limits are subject to change, and you’ll see an in-product notification when you hit a limit.

How will we know when we hit an additional usage spending limit?

You will get an in-product notification when you hit the additional usage spending limit.

Why can’t we purchase as much additional usage as we need without upgrading to a more expensive plan?

Additional usage spending limits are a temporary measure to mitigate abuse while we observe user behavior under the new billing model. We intend to pull these back in the coming weeks, at which point you will be able to purchase additional usage without upgrading your subscription.

Why do the spending limits vary? Why are some users treated differently?

Limits vary by plan tier and account signals, including usage patterns, billing history, and verification state, to protect service reliability and prevent abuse.

How exactly are you determining the amount of a user’s additional usage spending limit? What's the specific criteria?

We look at a variety of factors, including usage and payment history with GitHub. We can’t disclose specifics, as it would give bad actors more details to work with to circumvent our safeguards.

Will we always have to upgrade to a more expensive plan when we hit an additional usage spending limit?

No, we are working on solutions for allowing users to purchase as much additional usage as they need without upgrading to a more expensive plan. That said, our subscription tiers are designed to deliver more value as they increase in cost.

If I upgrade to a more expensive plan but then decide to downgrade, what does that look like? Are there any restrictions on when I can downgrade?

You can downgrade at any point. The downgrade will be a “delayed downgrade” where you will stay in the current, more expensive plan until your next billing cycle. Then you will be charged for the downgraded plan and the lower included usage will take effect the next calendar month.

Please note that if you upgrade from a Student plan to a paid plan, downgrading will take you to a Free plan and you will need to reapply for Student plan.

You said you’d loosen restrictions once usage-based billing is in effect, but sign-ups for new users are still paused. What’s going on?

We understand the frustration. Usage-based billing is now in effect, but we’ve needed additional time to put safeguards in place to ensure a reliable experience for all users. We plan to re-enable new sign-ups for all Copilot plans in the coming weeks.

What does it look like in practice to upgrade? Can you give an example of what the flow looks like?

Once you hit an additional usage spending limit, you can upgrade to the next plan to continue using Copilot’s full set of features (code completions and next edit suggestions remain unlimited for all paid plans). By upgrading, you will be paying a prorated amount of the next plan’s cost based on your billing cycle. Your accrued additional usage will be charged during your next billing cycle.

For example, once you hit the $29 additional usage spending limit on Copilot Pro and decide to upgrade, you will pay a prorated amount of the $39 Copilot Pro+ cost and receive $70 worth of AI Credits that are currently included with the Copilot Pro+ plan. On your next billing cycle, if you stay on Copilot Pro+, you will pay the full $39 for Copilot Pro+, the $29 in additional usage from Copilot Pro, and any additional usage from Copilot Pro+. A similar experience applies from Copilot Pro+ to Copilot Max with the corresponding costs.

When can I sign up for Copilot Max?

Copilot Max is available to existing Copilot Student, Pro, and Pro+ users that would like to upgrade for more included usage. We plan to enable Copilot Max sign-ups for new users when we re-enable sign-ups for all Copilot plans in the coming weeks.

Will you continue to offer flex AI Credits after the current promotion ends?

Flex allotments may vary from month to month. They’re designed to adapt as the economics of AI evolve, including model pricing, new models, and improvements in efficiency.

Why should I pay GitHub API prices when I can just go pay the model provider directly?

Copilot gives you a single subscription across every model family we support, including OpenAI, Anthropic, Google, xAI , with model availability varying by plan— without managing separate API keys, billing relationships, or rate limits with each provider. You get one bill, one set of usage controls, and access to all of it across VS Code, the Copilot CLI, the Copilot SDK, Copilot Code Review, and tools like OpenCode — without configuring each one separately.

For a lot of developers, the value isn't just "access to a model." It's that Copilot is deeply integrated into the tools they already use: pull requests, code review, the editor, the terminal. That integration work, including the context from your repo, your PR diff, your open files, and more, is what turns a model call into something more useful.

Will users continue to see the same level of restrictive rate limiting we’ve seen over the past few months?

We’ve loosened rate limits considerably, and users should not experience them with normal use.

Can I set up budgets at the user-level for my organization?

Yes, admins can now set a universal budget for users or override for specific sets of users. As users approach their budgets, admins will receive email notifications and can adjust budgets anytime from their billing settings. See our budget management documentation for details.

[EuSouVoce]

So the answer is basically: if you hit your limit too quickly, you should upgrade and pay more?

I get why GitHub moved to usage-based billing, but the numbers feel way off compared to the previous system.

I had only two short sessions in the VS Code extension, with maybe 2–4 prompts total, and somehow that already consumed 210.5 AI Credits/Tokens/Whatever.

Under the old model, that kind of usage would barely make a dent in my monthly quota. Now it feels like a handful of normal interactions can burn through a noticeable percentage of the month's allowance.

What worries me, and most of the users, isn't paying for usage. It's not knowing what a few everyday prompts are going to cost. If developers can't reasonably predict their consumption, it's hard to trust Copilot as a tool they'll use throughout the month.

Honestly, this feels like the kind of change that pushes people to look at other providers or even go directly to the model vendors instead of using Copilot as their main AI tool.

More transparency around how credits are consumed in real-world VS Code workflows would go a long way.
And remember... money doesn't grow on trees nor is it charcoal to be burned!

[rabbitao]

I made only about 7 requests, yet I consumed 48 units of my quota—this is simply unacceptable. Under the previous billing model, this would have consumed only about 1%.

[anuradhab008]

really fed-up. im unsubscribe my personal the pro account

[ferozsho]

Switched to Google AI. The 5-hour reset isn't ideal, but it's still good to have. Copilot has been a complete waste of money and time.

[AbramNel]

This change made it so much more expensive to do anything with it. I moved to just using chatgpt codex, it's only a little less user friendly, but you get so much more credits, and the models do a good job still.

[Fu-Jie]

My GitHub account is associated with Fu-Jie, and I have been a long-time annual subscriber.

My annual membership was originally set to expire in approximately two months. Earlier today (June 1, 2026), I clicked the "Refund" button in my account settings out of curiosity—not with the intention of immediately cancelling. I expected this action would lead to a confirmation page, not process the refund directly. However, the refund was executed immediately without any secondary confirmation prompt.

I believe this is a design flaw in the refund workflow. Any action affecting payment and subscription status should include an explicit confirmation step to prevent unintended consequences from accidental clicks or exploratory actions. The absence of this basic safeguard led to the unexpected cancellation of my membership.

To clarify: My request is not about renewal, but about the value of my current active contract. My existing membership had not yet expired and still had nearly two months of service remaining. Given recent changes to GitHub's billing policies, the terms and benefits under my current contract hold unique value during this remaining period. Once refunded and cancelled, renewing would not provide the same conditions.

Given the above, I request that my original membership status and expiration date be restored, and I strongly recommend that your team improve the refund process by adding a necessary confirmation step.

I look forward to your reply.

[prcris]

Eu também fiz isso no passado e ocorreu o mesmo, eles nem perguntam, te cancelam na hora. Muito louco isso.

Hi @Fu-Jie,

Thank you for writing in and explaining the situation.

We understand from what you shared that clicking the refund button processed immediately without a confirmation step, and that this was not your intent. We appreciate you flagging this as a UX concern. I've noted this feedback and shared with internal teams for investigation.

Regarding restoring your previous subscription: this falls outside what I'm able to action directly. We cannot resolve account related issues in this public forum, and I'm sharing this information with the appropriate team to review whether your original subscription and expiration date can be reinstated.

I'll follow up once I have more information. Thank you for your patience.

[chalkplum]

one afternoon pro+ limit 85%

[feuergott]

i havent received any additional credits the whole thing is disgusting, even there so called MAX plan looks a joke at 2.9x pro+
You DID get your additinal credits. Those 7000 credits ARE the promotional credits. As you can see in the billing dashboard, those are at 50% discount. Prepare for getting half the credits in the (near) future!

[sasengineering2025-a11y]

Yes here same 2 days work 7K credits burned.. god bye GitHub.. you are not more payable, and its close to robbery! besides Copilot keeps you busy after restart to fix the VS again.. +its a shame thins so many developers here put things for free with all there know how, which get robed too. Years a go GIT was relay good now its just a massive rib-off...very sad!

[fsoft72]

I was a Github Copilot beta tester, then one of the first paying for the service. I was an evangelist of Github Copilot. Today I canceled my subscription. I am switching for Antigravity and Claude, and I am also considering testing how good the new Cursor is.

[bnevolve]

I upgraded to MAX and started to restructure some docs in my project and after that I did a few task for some implementations - I ended up spending all my MAX credits under 24 hours, this is crazy....

I was working heavily almost every day last month on Pro+ and a budget on $100 - no problem...

Same here the costs are exploding in just a couple of days. It is expected that some raises are necessary but these exceptional higher use of the Agents models and costs are way out of balance

[ElettroAle]

The new pricing model is outright robbery!!! I've already canceled my GitHub Copilot plan subscription!!

did the same

[dobrowolskijaroslaw1-sys]

Subject: Large day-to-day difference in Copilot usage for similar work

Hello,

I am reporting a large difference in Copilot usage between two recent sessions on the same repository. On 2026-05-31, I was able to work through a long session and the usage remained within a range that felt predictable. On 2026-06-01, a session that felt comparable or smaller in visible scope consumed much more of my quota. From my point of view the difference is easy to verify and does not match the apparent amount or difficulty of the work performed. I understand that internal context, tool payload, or other service-side factors may exist, but the current accounting behavior is not transparent enough for planning real project work. Please clarify whether the accounting model changed or whether hidden context growth is being billed in a way that users cannot see. I can provide screenshots, approximate session times, and a short written comparison of the two sessions. 8 hours not heavy work...

Regards,
Very disapointed...

[lmerlas]

It is disappointing to see the response to these concerns shift into a 'blame game' where users are essentially told they have a 'skill issue' for not optimizing tokens, rather than GitHub providing a transparent, predictable estimation model before requests are run.As it stands, the 'integration' advantage doesn't justify the cost or the lack of predictability.

For my personal work, I’m moving to Codex and exploring other providers immediately.
For my team, the economics of a $0.5M internal deployment for 100 seats now far outweigh the headache of usage-based billing, privacy concerns, and the nightmare of renegotiating third-party NDAs.

[LTSarc]

It is disappointing to see the response to these concerns shift into a 'blame game' where users are essentially told they have a 'skill issue' for not optimizing tokens, rather than GitHub providing a transparent, predictable estimation model before requests are run.As it stands, the 'integration' advantage doesn't justify the cost or the lack of predictability.

For my personal work, I’m moving to Codex and exploring other providers immediately. For my team, the economics of a $0.5M internal deployment for 100 seats now far outweigh the headache of usage-based billing, privacy concerns, and the nightmare of renegotiating third-party NDAs.

I hope you realize that Codex will do the same as well if they get the huge amount of transferred users. The old rates were simply massively underwater financially and MS has drawn a line.

[BrandonTaylorDev]

It is disappointing to see the response to these concerns shift into a 'blame game' where users are essentially told they have a 'skill issue' for not optimizing tokens, rather than GitHub providing a transparent, predictable estimation model before requests are run.As it stands, the 'integration' advantage doesn't justify the cost or the lack of predictability.
For my personal work, I’m moving to Codex and exploring other providers immediately. For my team, the economics of a $0.5M internal deployment for 100 seats now far outweigh the headache of usage-based billing, privacy concerns, and the nightmare of renegotiating third-party NDAs.

I hope you realize that Codex will do the same as well if they get the huge amount of transferred users. The old rates were simply massively underwater financially and MS has drawn a line.

That is a "them" issue. They should not have made a product with high value, low cost, tie it into systems that a majority of developers use, push it at every turn, then yank the proverbial rug out saying "Oops! Too expensive."

[LTSarc]

It is disappointing to see the response to these concerns shift into a 'blame game' where users are essentially told they have a 'skill issue' for not optimizing tokens, rather than GitHub providing a transparent, predictable estimation model before requests are run.As it stands, the 'integration' advantage doesn't justify the cost or the lack of predictability.
For my personal work, I’m moving to Codex and exploring other providers immediately. For my team, the economics of a $0.5M internal deployment for 100 seats now far outweigh the headache of usage-based billing, privacy concerns, and the nightmare of renegotiating third-party NDAs.

I hope you realize that Codex will do the same as well if they get the huge amount of transferred users. The old rates were simply massively underwater financially and MS has drawn a line.

That is a "them" issue. They should not have made a product with high value, low cost, tie it into systems that a majority of developers use, push it at every turn, then yank the proverbial rug out saying "Oops! Too expensive."

I mean, it's immoral but also entirely predictable and a classic business strategy. You were effectively all on a discounted preview. If you look at API costs on any frontier model, it is obvious MS was eating over 90% subsidy here. MS may live on a mountain of money but the board would never approve of this forever.

[ThePlenkov]

Dumbest decision ever....

[YuriySamorodov]

Guess what? It's called effective management, hah

[DanielRuf]

Hm, wondering if there will be similar changes for the many Copilot variants in M365 and general Microsoft products in the future.

@DanielRuf 👋🏻 Thanks for the question. Our team here supports GitHub Copilot specifically, and we don't have visibility into billing plans for Microsoft 365 Copilot or other Microsoft products.

For questions about M365 Copilot, Microsoft's product and support channels would be the best resource.

I am assuming this was posted in jest, but answering with added context for you or any other users where the info may be helpful.

[LESdylan]

I had faith in you.. Honestly, I regret not taking the refund offer you made. I genuinely think you're going to lose at least 80% of your client base. This is so ridiculous. It's far better to pay in real money and not by token. The base usage is ridiculously low.

[YuriySamorodov]

I regret not taking the refund offer you made

I hear you.. Sound like a trap indeed. So sad and uneven

[gyzrich]

The new pricing model is outright robbery!!! I've already canceled my GitHub Copilot plan subscription!!

[VoxSecuritatis]

I think it is more of a pendulum; yes, intially a lot of people will leave, but I have a feeling they will come back... eventually.

[tofanelli]

I think it is more of a pendulum; yes, intially a lot of people will leave, but I have a feeling they will come back... eventually.

most likely they won't for a fairly simple reason... you have set your whole workflow on copilot, adjusted every single part to your needs and you had to, forcefully, to change to another app, another workflow, another settings and so on.... once you got used to that new one, you won't be coming back even if the older one is better just because you are tired to going back and forth... you most likely won't be doing any of this. you will stick to the new one and start adjusting yourself

[jakteixeiraf]

I'm using VS Code with GitHub Copilot and I chose LLM GPT 5.4 Mini, but the platform is deducting the cost of Gemini 3.5 Flash, and because of that, my credits are disappearing without me using them.

[jakteixeiraf]

I'm currently using SONET 4.6 and it's charging me for GPT 5.4 mini, are you kidding me?

@jakteixeiraf 👋🏻 Hi there,

Thank you for reporting this — I want to make sure we understand what's happening.

To clarify: you selected GPT 5.4 Mini in VS Code Chat, but your usage report shows charges attributed to Gemini 3.5 Flash. Is that correct?

A couple of things to be aware of:

• Some Copilot features (e.g., code completions, agent sub-tasks, or background operations) may use different models than the one selected in your Chat panel. These are separate from your explicit Chat model choice.
• Your detailed usage CSV will show exactly which model was used for each request.

Could you please:

1. Download your usage CSV from Settings → Billing and licensing → Metered usage → Copilot
2. Confirm whether the Gemini 3.5 Flash charges correspond to Chat requests (where you selected GPT 5.4 Mini) or to other features like completions or background operations

If the CSV confirms you're being charged for a model you didn't select for those specific interactions, please share that detail and we'll investigate further.

[ronan-fesselier]

the admin itself is a f*****g bot :)

[Mieekaserra]

How in the hell can 1 prompt which last month would barly cost a cent, now take up $4.51.

this is ununsable and needs rolling back.

[suspex0]

Yes one small UI bug fix 1,50$ lmao

[Plushie905]

Imagine , 2 calls , 550 credits so if i'm in Pro plan = 3 commands -_-

[gyzrich]

The new pricing model is outright robbery!!! I've already canceled my GitHub Copilot plan subscription!!

[hikerID]

Check out the input/output costs for some of the agents, it's insane. Claude Sonnet is 300 input, 1500 output, 30 for cache.

[SuperManito]

So why has the Pro plan reduced its usage limit by almost 100 times compared to before? I also discovered that the GPT-5 mini model has become less intelligent.

👋🏻 @SuperManito Hi there, Thank you for flagging this.

Regarding the usage limit change:

The billing model has shifted from counting flat requests to measuring actual token consumption (expressed as AI credits). These are different units, so a direct numerical comparison between the old request count and the new credit balance isn't apples-to-apples.

Under the previous model, every request counted as "1" regardless of complexity. Under usage-based billing, a simple one-line completion costs very little, while a large multi-file generation costs more, reflecting actual compute used.

Importantly:

• Code completions (autocomplete) remain unlimited and are not charged per token
• Your per-seat subscription price has not increased
• You can monitor exactly where credits are going in Settings → Billing and licensing → Metered usage → Copilot

Regarding model quality:

Model behavior feedback is best surfaced through the 👍/👎 buttons on Copilot responses in VS Code - this goes directly to the team responsible for model quality. If you're seeing a consistent degradation, that feedback is valuable.

For managing your usage:

• https://docs.github.com/en/copilot/tutorials/budgets/getting-started-with-budget-controls

[MissyGH]

You should have tested this on people instead of using AI. Most of us are willing to pay $10/month but not $100 or more per month. The mystery billing is problematic because we can use up a quarter of our budget on the 3rd and end up with nothing left by the 10th. Then we are stuck trying to decipher the zillion lines of AI generated code in order to debug problems created with the assistance of AI.

I hope you will consider how this impacts your customers.

[LTSarc]

You should have tested this on people instead of using AI. Most of us are willing to pay $10/month but not $100 or more per month. The mystery billing is problematic because we can use up a quarter of our budget on the 3rd and end up with nothing left by the 10th. Then we are stuck trying to decipher the zillion lines of AI generated code in order to debug problems created with the assistance of AI.

I hope you will consider how this impacts your customers.

Brother you can't even run hundreds of thousands of tokens on a middling, non-Agentic model for $10/month.

As someone who has used independent providers and hosted my own locally, frontier models are staggeringly expensive. Power users were eating up thousands and thousands of dollars in model use.

[neohemp]

You should have tested this on people instead of using AI. Most of us are willing to pay $10/month but not $100 or more per month. The mystery billing is problematic because we can use up a quarter of our budget on the 3rd and end up with nothing left by the 10th. Then we are stuck trying to decipher the zillion lines of AI generated code in order to debug problems created with the assistance of AI.

I hope you will consider how this impacts your customers.

Brother you can't even run hundreds of thousands of tokens on a middling, non-Agentic model for $10/month.

As someone who has used independent providers and hosted my own locally, frontier models are staggeringly expensive. Power users were eating up thousands and thousands of dollars in model use.

Except they did do that
And they took people's money up front for a year and then backed out on it

[LTSarc]

You should have tested this on people instead of using AI. Most of us are willing to pay $10/month but not $100 or more per month. The mystery billing is problematic because we can use up a quarter of our budget on the 3rd and end up with nothing left by the 10th. Then we are stuck trying to decipher the zillion lines of AI generated code in order to debug problems created with the assistance of AI.
I hope you will consider how this impacts your customers.

Brother you can't even run hundreds of thousands of tokens on a middling, non-Agentic model for $10/month.
As someone who has used independent providers and hosted my own locally, frontier models are staggeringly expensive. Power users were eating up thousands and thousands of dollars in model use.

Except they did do that And they took people's money up front for a year and then backed out on it

You still got the year you paid for, they're just opting out of offering it going forward. I'm not trying to defend MS' actions here, just the logic behind them and how predictable it was. Have you seen what OAI or Anthropic API key rates are?

[CCarrMcMahon]

In case anyone is curious on the value comparison, this is what I am seeing on the $10 plan using GPT-5.4:

• Old Version:
  • 300 premium requests (3 messages = 1% usage)
• New Version:
  • 50 premium requests (3 messages = 6% usage)
  • 1000 credits (3 messages ~= 100 credits = 10% usage)

The 100 credits is an upper estimate of what I have been seeing if you assume larger prompts. I have seen as low as 10 credits and as high as 60 so far today. On average I seem to hover around 33 credits per prompt.

@CCarrMcMahon 👋🏻

You're right that the units have changed: the old model counted "requests" regardless of complexity, while usage-based billing measures actual token consumption. For users who frequently use higher-cost models like GPT-5.4 with larger context, the per-interaction cost is higher than a flat request count implied.

A few things that may help manage your credits:

• Model choice matters significantly. GPT-5 mini ($0.25/1M input, $2.00/1M output) is substantially cheaper than higher-tier models. Choosing a model appropriate to the task complexity can stretch your credits considerably.
• Code completions remain unlimited and are not charged per token on all paid plans.
• Caching helps over time. Repeated context (e.g., the same files across a session) gets cached automatically at up to 90% lower cost.
• Shorter, focused prompts with only necessary context reduce input tokens.

You can see exactly what each request costs in your usage CSV: Settings → Billing and licensing → Metered usage → Copilot.

For budget controls and notifications:
https://docs.github.com/en/copilot/tutorials/budgets/getting-started-with-budget-controls

[TMattes]

@CCarrMcMahon 👋🏻

You're right that the units have changed: the old model counted "requests" regardless of complexity, while usage-based billing measures actual token consumption. For users who frequently use higher-cost models like GPT-5.4 with larger context, the per-interaction cost is higher than a flat request count implied.

A few things that may help manage your credits:

• Model choice matters significantly. GPT-5 mini ($0.25/1M input, $2.00/1M output) is substantially cheaper than higher-tier models. Choosing a model appropriate to the task complexity can stretch your credits considerably.
• Code completions remain unlimited and are not charged per token on all paid plans.
• Caching helps over time. Repeated context (e.g., the same files across a session) gets cached automatically at up to 90% lower cost.
• Shorter, focused prompts with only necessary context reduce input tokens.

You can see exactly what each request costs in your usage CSV: Settings → Billing and licensing → Metered usage → Copilot.

For budget controls and notifications: https://docs.github.com/en/copilot/tutorials/budgets/getting-started-with-budget-controls

@github Community Admin
I also see a major issue with the output cost, which is significantly higher than the input cost: users have no control over the number of tokens the model generates. While users can keep their input short and focused, every LLM I’ve seen so far tends to be far too verbose, which consumes a large number of tokens.

And what if the output is completely useless? Users still have to pay for it—and on top of that, they then have to spend additional time fixing the mistakes in an often lengthy and expensive response.

Furthermore, the creativity is completely lost, if you constantly have to monitor your token consumption and then realize, that your tokens are gone, before you even reach a viable result.

You really should overthink that, otherwise you will loose a lot of customers...

[lelacag]

Join the $10Club :) I am not going to change it at all. I just want to see where it goes. I hope they change their way of billing and continue to support the developers.

[chalkplum]

GPT-5 mini lol

[aidannairn]

In the previous thread about Usage-based Billing it seemed that a common gripe people had was that unused tokens do not roll-over into the next month. To my knowledge, the topic was never addressed by GitHub. Are GitHub able to disclose the factors contributing to this decision?

While GitHub make a valid point regarding the convenience of having one bill and one set of usage controls—I’ve opted to stick to Pro+, burn through my allocated credit in a week, and then pivot to using OpenRouter for the remainder of the month. OpenRouter offers a similar set of advantages that Copilot has over other providers. It can be used within the same VS Code interface. Plus it has more models and credit rolls-over for up to a year.

[neohemp]

Thank you for the feedback.

To confirm: included AI credits reset each billing cycle and do not roll over to the following month. This applies to all plans.

Regarding the reasoning behind this decision: I don't have product strategy context to share on that front, but feature requests and feedback like this are across the GitHub Community Discussions are tracked by the product teams, and we report on what we're hearing from users regularly, so continue to share here.

We appreciate you sharing the comparison and your workflow. That kind of feedback is valuable for the team to understand how users are adapting.

Let them know I cancelled my office365 too because of your greed

[oschuda]

Open Router can't change and edit files, it is useless for me, as I am not a programmer, I am a systems admin.

[webbugt]

Open Router can't change and edit files, it is useless for me, as I am not a programmer, I am a systems admin.

Try opencode (has a web ui and vscode plugin) with openrouter as provider. Add some specialized agents and you're off. You can even mix and match providers as you wish. Basically supports all of them since if not supported directly you can input any openAI standard API as a custom provider.
There's also a plugin to use existing copilot instructions inside opencode

[aidannairn]

Open Router can't change and edit files, it is useless for me, as I am not a programmer, I am a systems admin.

That may have been true some time ago. I can say with absolute certainty that using OpenRouter within VS Code is as seamless as using Copilot’s own models.

If you ask Google today, the AI response will likely give you some outdated info about requiring an extension for it to work. So I understand the skepticism.

Creating an account is free. Why not give it a try using one of their free models? That’s what I did last month when I was weighing out my options.

[nikolaydubina]

Open Router can't change and edit files

what do you mean? get token. connect in VSCode. all AIs that support tools connected throuhg standard VSCode can use 100% of tools there. vision, files, everything. from gemini class models to deepseek or anything, local or cloud, does not matter. as long as they speak OpenAI API.

[neohemp]

Posting here so I can still subscribe to the chaos

[alesanGreat]

I've been a paying Copilot user and this new credits system is simply unacceptable. In less than 5 hours of normal work today — the very first day of the new billing — I've already burned through 50% of my monthly plan. This is not a pricing change, this is a bait and switch. We paid for a tool, and overnight it became something we can't afford to use the way we always have. I'm canceling my subscription unless this is reverted or seriously reconsidered.

[iserje]

I use the same model GPT-5.4 before and after, so please don't blame the model selection. And now in 2 days, I'm almost out of credits meaning value of Pro+ license is pretty much 0.

[fishwang001]

It's not just faster. It's rocket shoot...

[kerwin-git-2020]

gpt5.4 is previously 1X, now 6X.

[mhzaki1]

this is fraustrating , i notice in 4 prompt of sample question it used 64% of my budget , what is going on? i tried to contact some human support of github more than 10 times, but nothing

[e-nikulov]

What?? Github, but you just said to use short question, one topic per question to keep small context and now you recommended to repeat content. You mean your BIG but cached context. Great. Just continue spent money.

[jarrocha]

We are not the main target customer anymore. If you read the news, you can see that large companies are paying millions for this type of service, they don't need users like us.

[neohemp]

We are not the main target customer anymore. If you read the news, you can see that large companies are paying millions for this type of service, they don't need users like us.

You're right Microsoft is trusted by a lot of companies but I don't think they realize when you do customers like this they remember it

[sl2365]

Time to ditch Microslop then and go with linux!

[Wakko13]

I would like to know why you're charging for use on days I haven't used Copilot (Today). I've moved to using Claude Code and GPT Codex since this billing mess isn't fixed.

Is this running some kind of sync from VSCode against Copilot and charging back to the users?

[Wakko13]

I've now disabled the extensions for Copilot on both VSCode install locations. If I see you charge again, I'll be reporting this as theft.

[cachebag]

having the same issue. i have intentionally avoided copilot at work and i am still being charged.

Thanks for sharing the details @Wakko13. Community Discussions can’t review account-specific billing or usage records, so we can’t confirm what caused the increase here. If you believe the usage shown for your account or business seat is incorrect, please contact GitHub Support so they can review the relevant account and billing details.

When contacting Support, please include:

• the organization or billing entity where you see the usage increase,
• the date range and timezone,
• the usage view or screenshot you’re referencing,
• what you expected the usage to be,
• any details about Copilot surfaces that were enabled or disabled during that time.

If you already have a Support ticket open, please continue working through that ticket and wait for Support’s response. We’re not able to escalate or resolve Support tickets from Community Discussions.

[kindsee]

Hi all,
I am extremely disappointed with GitHub Copilot’s new credit system.

With the same usage as always, my credits are now depleted within minutes. On top of that, the system now goes in absurd circles, giving unnecessarily long or off‑track responses — and with this new billing philosophy, it’s obvious why: every extra detour burns through credits in just a few hours.

This makes the service completely unsustainable and, honestly, it makes me feel scammed.

I’m not going to keep paying for a product that used to work well and now punishes users for working.

I will be canceling my subscription and looking for more transparent and reasonable alternatives.

I hope GitHub reconsiders this model, because right now you’re just pushing people away from the platform.

[cardonator]

I don't like this but I'm wondering what people think GitHub's options are here. I am someone who in April spend $20 in overages on the Pro plan but my actual cost to GitHub was over $2,000 (I wasn't trying to game anything, btw, it's just how my usage worked out).

They might have more options with their own model but people would still complain about that being worse than a everyone else. They are passing these costs straight through, and every model providers has tightened their quota based usage for individuals, and switched enterprises to token based billing.

So while this change is super annoying, I really don't see how GitHub can be expected to keep funding the gravy train while our usage costs them 2-100x what we are paying.

[neohemp]

I don't like this but I'm wondering what people think GitHub's options are here. I am someone who in April spend $20 in overages on the Pro plan but my actual cost to GitHub was over $2,000 (I wasn't trying to game anything, btw, it's just how my usage worked out).

They might have more options with their own model but people would still complain about that being worse than a everyone else. They are passing these costs straight through, and every model providers has tightened their quota based usage for individuals, and switched enterprises to token based billing.

So while this change is super annoying, I really don't see how GitHub can be expected to keep funding the gravy train while our usage costs them 2-100x what we are paying.

Well it's not like we told them to do that. At the VERY least they could have continued the yearly accounts until they expired (and I wasn't a yearly subscriber). It seems in poor faith to just say nope we're not giving you what you actually agreed upon.

[constantin-minea]

We probably would understand it if they wouldn't charge 4-5 times more credits than OpenAI or Anthropic. Plus the whole other stuff that uses credits in the background which they haven't thought of. So no, it is poor implementation, and way more expensive than it supposed to be. Canceled, went to OpenAI, Codex 5.3 is charged at 43.75 | 4.375 | 350 and NOT 175 | 75 | 1400. Mindblowing they thought this will fly. Ha Ha Ha.

[Posuza]

I his that just for using around 15 minute after my ai is check for project production pipe line

[Wakko13]

If you have overages set, you might want to zero them out until these billing/usage issues are resolved.

[tinezivic]

I think GitHub was probably willing to subsidize the high costs earlier because advanced users were effectively helping them develop and improve Copilot through heavy real-world usage. Now that the product is more mature, it looks like they are moving those costs back to the users with usage-based billing.

[agtMorpheus]

My limit changed from one month, which matched my normal usage flow, to just three days. Now I’m expected to wait 27 days before using it again.

Great move. I’ve canceled it.

[fershibli]

A single day of use and it used 55.9% of my monthly premium requests.
This is absurd, the previous month I've used less than this for the ENTIRE MONTH - using the same amount every single business day.

You GitHub and Microsoft, you're being the WORST for your own users.
This needs to be rolled back and people need to get their tokens back.

[MartinovEm]

Keeping your Copilot subscription after June 1st is madness, and their policy is a rip-off!

I highly recommend installing the Claude Code extension in VSCode (I’ve been using Copilot there until now). For €90, you get Claude Opus 4.8 with 1MB of context, which is... a real, REAL beast! And it is your default model there :) Even Claude Sonnet 4.6 in Claude Code is much, much more capable than Sonnet 4.6 in Copilot. Sonnet 4.6 in Claude Code is perhaps as capable as the previous Opus 4.6 in Copilot, but it might even be better than that.

After 2 days of coding, averaging 6-7 hours a day, I’ve used up less than 15% of my weekly limit.

GH has simply shot itself in the foot with this financial “policy.” That's it.

[dhovang]

I did the same. Hit my monthly budget this morning after 3 days. I've never hit the monthly limit before. After reading up I switched to Claude and cancelled my CowPilot seat. The switch took about 20 minutes, including signup. Claude has much better transparency on cost and it's much cheaper than CowPilot. Also, it seems like its reasoning performance is better.

Good bye CowPilot.

[juanc-tiv]

Keeping your Copilot subscription after June 1st is madness, and their policy is a rip-off!

I highly recommend installing the Claude Code extension in VSCode (I’ve been using Copilot there until now). For €90, you get Claude Opus 4.8 with 1MB of context, which is... a real, REAL beast! And it is your default model there :) Even Claude Sonnet 4.6 in Claude Code is much, much more capable than Sonnet 4.6 in Copilot. Sonnet 4.6 in Claude Code is perhaps as capable as the previous Opus 4.6 in Copilot, but it might even be better than that.

After 2 days of coding, averaging 6-7 hours a day, I’ve used up less than 15% of my weekly limit.

GH has simply shot itself in the foot with this financial “policy.” That's it.

I was forced to go with the pro+, the $10 pro was more than enough before june 1st. so now its like $40 and tokens are going pretty fast, but paying anything $50 is not good either, $90-$100 for ai? thats insane i rather go ai-less lol

[Original-Iluster]

The power of a well-structured agent app + efficient cache invalidation—or, conversely, refraining from it when it's unnecessary—makes this possible:

And even if DeepSeek ever starts to feel a bit too risky for me, there’s always Mistral—goodbye, GitHub Copilot!

[MarkJB]

I hit 45% yesterday after about ~6hrs of usage over 3 days.

And this is with careful "Ask" and "Plan" usage before switching to "Agent" and using "Auto" for the 10% discount... 😞

[belaszalontai]

GitHub Copilot is DEAD? Do This After June 1st!
sounds reasonable: https://youtu.be/_uUAdIcpxQE

[julian-donald]

My forecast cost has just gone up 100x and I've reluctantly canceled my subscription which is a real shame as the product was getting very good. I now have the cost of moving to an alternative but over the next few months it will still work out significantly cheaper than staying with ms. Unfortunately for ms when customers invest in moving to a new supplier it's highly unlikely they will come back. Will be fascinating to see if ms make the same revenue off a much smaller number of customers paying significantly higher price?

[icodeintx]

DeepSeek is working perfect for me, if not better PLUS I'm not being robbed.

[frozenprakash]

Yes true, copilot is becoming completely useless after usage based billing, however I optimize my context & model, it's still consuming credits like crazy!

I had moved to openrouter.ai and the amount of models it provides and cost is extremely low, I could easily survive with same 20$ for entire month, if I am using the models carefully.