Home / Guides / Copilot Rate Limits
GitHub Copilot Rate Limits Explained
Everything you need to know about GitHub Copilot rate limits - Free vs Pro vs Business allowances, the AI Credits that replaced the premium request multiplier, and what happens when the allowance runs out.
Last updated October 2026 · By Soren Starck
What Are GitHub Copilot Rate Limits?
GitHub Copilot enforces two distinct limits depending on your plan. The Free plan caps monthly chat and completion counts. Pro, Pro+, Max, Business, and Enterprise plans remove the completions cap but meter chat, agent mode and other model calls against a monthly allowance of AI Credits (1 credit = $0.01), billed at each model's per-token rate.
The Plans at a Glance
| Plan | Code completions | Chat / month | AI credits / month |
|---|---|---|---|
| Free | 2,000 / month | 50 | An allowance (auto model only) |
| Pro ($10) | Unlimited | Uses credits | 1,500 ($15) |
| Pro+ ($39) | Unlimited | Uses credits | 7,000 ($70) |
| Max ($100) | Unlimited | Uses credits | 20,000 ($200) |
| Business ($19) / Enterprise ($39) per seat | Unlimited | Uses credits | 1,900 / 3,900 per user, pooled |
Once your monthly AI Credits allowance is gone, Copilot does not quietly switch you to a weaker model. GitHub gives you three choices: wait for the monthly reset, keep working on paid usage within a dollar budget you set, or move to a cheaper model yourself. Code completions keep working either way, because they use no credits on paid plans.
How AI Credits Are Charged
Not every request costs the same. Since June 1, 2026, GitHub bills by the model and the number of tokens, at the rates in its models and pricing table:
| Model / Task | Per 1M tokens (input / output) | Notes |
|---|---|---|
| Code completions | Not billed | Unlimited on paid plans, no credits used |
| GPT-5 mini | $0.25 / $2.00 | Lightweight |
| Claude Sonnet 5.5 / GPT-6.1 Sol | $2.00 / $10.00 | Versatile and powerful models |
| Claude Opus 5.5 | $4.00 / $20.00 | Heavy reasoning model |
| Agent / multi-step | Sum of every call | One ask, many model calls |
On Pro (1,500 credits, or $15, a month), an agent task that sends 1M uncached input tokens to Claude Opus 5.5 and gets 100K back costs $6, or 600 credits. Two of those a week and your allowance is gone before you notice.
Annual Copilot Pro and Pro+ subscribers who stayed on the legacy request-based billing still spend premium requests with model multipliers until their plan ends; GitHub lists those in its legacy multiplier table.
What Happens When You Hit the Limit
- Wait for the reset: your included allowance resets every month, and Copilot settings show the reset date
- Pay as you go: set a dollar budget and Copilot continues; credits draw down at $0.01 each, so a $10 budget covers 1,000 credits
- Switch to a lighter model: lightweight models use fewer credits per interaction and stretch what is left
- Business and Enterprise: admins set the limits; if they allow no paid usage, Copilot pauses until the next cycle
GitHub sends alerts at 75%, 90% and 100% of a budget you configure, and code completions stay unlimited on paid plans (GitHub Copilot plans FAQ).
SessionWatcher
Copilot credits run out mid-month.
SessionWatcher shows the burn.
Native app for the Mac menu bar and the Windows taskbar. Track Claude, Codex and 20 more tools' usage, costs, and rate limits in real time.
“Fast, simple, and does exactly what it should. Definitely worth it.”
@nicojerome on GitHub
macOS 14+. Free for 7 days, no credit card.
How to Monitor Copilot Rate Limits
SessionWatcher for Copilot is a native app for the Mac menu bar and the Windows taskbar, built for exactly this. It tracks:
- Premium request count against your plan's monthly cap
- Burn rate - premium requests per day, projected to month-end
- Model breakdown - which premium models are eating your quota
- Billing-cycle countdown - exact day/time of quota refresh
- Desktop notifications - at 70% and 90% by default, so you can pace
Tips for Managing Your AI Credits
- Reserve Opus for hard problems - Claude Opus 5.5 costs twice as much per token as Claude Sonnet 5.5.
- Lean on code completions for routine autocomplete - on paid plans they don't touch your AI Credits at all.
- Track burn rate, not just totals - 750 credits gone on day 5 means Pro runs dry by day 10.
- Time agent runs deliberately - multi-step agent tasks are the biggest quota eaters.
- Set notifications - SessionWatcher pings you before your AI Credits allowance runs out.
Frequently Asked Questions
What are GitHub Copilot rate limits?
Free plans cap at 50 chats and 2,000 completions/month. Paid plans bill chat and agent work in AI Credits at per-token model rates: Pro includes 1,500 credits a month, Pro+ 7,000, Max 20,000, Business 1,900 per user and Enterprise 3,900 per user. Multipliers now apply only to legacy annual Pro and Pro+ plans.
How long until my Copilot quota resets?
The AI Credits allowance resets at 00:00 UTC on the 1st of each calendar month, not on your billing date. SessionWatcher tracks the exact countdown for you.
How do I avoid hitting Copilot rate limits?
Monitor in real time with SessionWatcher. Copilot tracking is part of Pro ($59 one-time, or $24/year, covering all supported tools). It shows current usage, projected burn rate, and notifies you at 80% so you can pace your premium-model work.