GitHub Copilot Free Plan Explained: Limits, Models and Upgrades

GitHub Copilot's free tier gives you 2,000 code completions and 50 chat requests per month (including Copilot Edits), auto-selected AI models only, and basic agent mode in supported editors — but no cloud coding agent, no full pull request reviews, and no manual model picking. If you are coding daily, pairing heavily with chat, or want to assign background tasks to an agent, you will hit those limits fast and should look at the $10/month Pro plan instead.

What Is GitHub Copilot Free, Exactly?

GitHub Copilot Free is the no-cost tier of GitHub's AI coding assistant, aimed at individual developers who do not already get Copilot through an employer, school, or open-source maintainer program. It requires only a free github.com account — no credit card, no trial clock. According to GitHub's own Copilot plans page, Free gives you "limited access to a selection of Copilot features" so you can try AI-assisted coding before deciding whether to pay.

Under the hood, Free runs on the same product as the paid tiers — the same chat panel, the same inline completions, the same agent mode toggle in VS Code, Visual Studio, JetBrains IDEs, and Copilot CLI — just with firmer caps and fewer model choices. It is not a stripped-down demo; it is the full Copilot experience with a monthly meter attached.

The Actual Free-Tier Limits: Completions, Chat and Agent Mode

GitHub states the Free plan's numeric caps plainly on its plans page: "GitHub Copilot Free users are limited to 2000 completions and 50 chat requests (including Copilot Edits)." Those two numbers reset every calendar month. A couple of details matter here:

• 2,000 completions/month covers inline code suggestions and next-edit suggestions as you type — this is the ghost-text autocomplete most people associate with Copilot.
• 50 chat requests/month covers Copilot Chat conversations and Copilot Edits (multi-file edits driven by chat). A single back-and-forth chat session can burn through several of those 50 requests quickly if you ask follow-up questions.
• Agent mode is listed as included for Free in GitHub's feature comparison table, but GitHub's own documentation on individual Copilot plans separately describes agent support on Free as "Limited," and the cloud/background coding agent — the one you assign an issue to and come back for a finished pull request — is explicitly not included on Free.
• Model selection on Free is auto-only: Copilot picks the model for you behind the scenes, and you cannot manually switch to, say, a specific Claude or GPT model.
• Code review on Free is limited to "Review selection" inside VS Code — the full pull-request-level AI code review feature is a paid-plan feature.

Once you exhaust either the completions or chat allowance in a given month, inline suggestions and chat simply stop working until the next reset — Free does not let you pay for overage the way paid plans do.

Which AI Models Does Copilot Free Actually Use?

This is the part free users ask about most, and it is also where GitHub is least specific. The supported models reference says plainly that "the models available to you depend on your Copilot plan and where you're using Copilot," and GitHub's billing documentation confirms that Free is restricted to auto model selection — the system routes your request to whichever underlying model (from GitHub's roster of OpenAI, Anthropic, Google, Microsoft, xAI and other partner models) it judges appropriate for that request, balancing quality, latency and cost on GitHub's side.

What you don't get on Free is a model picker. Pro and above let you manually choose from a list of available models for chat and agent sessions, and Pro+ and the higher Max tier add access to additional premium-tier models (GitHub cites Claude's Opus-class models as an example of a "premium model" gated above Pro). If a specific model's behavior or cost profile matters to your work — for example, if you've read about a new model like the one covered in our rollout of Gemini 4 Argon inside Copilot — Free won't let you request it directly; you're at the mercy of auto-selection.

How GitHub's "AI Credits" System Changed the Math

In mid-2026, GitHub overhauled Copilot billing. As GitHub announced on its engineering blog, every Copilot plan moved from the old flat "premium request" quota system to usage-based GitHub AI Credits, effective June 1, 2026. The mechanics, per GitHub's models and pricing documentation:

• 1 AI Credit = $0.01 USD.
• Credits are consumed based on the actual input, output and cached tokens a request uses, priced per-model — a short chat question on a small model costs little; a long agent session on a larger model costs more.
• Code completions and next-edit suggestions never consume credits — only chat, Copilot Edits, agent mode and other "agentic" interactions do.
• Paid individual plans get a monthly credit allotment split into a base amount and a "flex" amount, and can buy additional credits if they run out. Business and Enterprise get pooled, per-seat allotments administrators can cap.

Here is the catch for Free users specifically: GitHub's plan comparison still describes Free's chat ceiling in the older, simpler language — the flat 50-requests-per-month cap — rather than quoting a specific AI Credits number the way it does for Pro ($15 total credits: $10 base + $5 flex), Pro+ ($70 total: $39 base + $31 flex) and the newer Max tier ($200 total: $100 base + $100 flex). GitHub's own pages describe Free's allowance only as "limited chat and agent usage" without publishing an exact credit figure, so treat the 50-chat-request number as the operative, documented cap for Free rather than trying to convert it into a credit amount yourself.

If annual billing sounds familiar from older Pro/Pro+ research you've read, that's because GitHub let existing annual subscribers keep the legacy premium-request system (300/month for Pro, 1,500/month for Pro+) until their current annual term ends, per GitHub's request-management documentation — but every new signup, every monthly subscriber, and Free itself now operates under the credits framework described above.

Free vs Pro vs Pro+ vs Business vs Enterprise: Full Comparison

The table below pulls together GitHub's published pricing and feature pages for every current Copilot tier.

Figures above come from GitHub's plans page and GitHub's individual-plans documentation. GitHub notes seat prices for Business and Enterprise did not change in the June 2026 billing shift — only how usage above completions is metered did.

What Business and Enterprise Add on Top

Business and Enterprise are organization-level plans, bought per seat by an admin rather than by an individual. Beyond the larger pooled AI Credit allotments in the table above, GitHub's plans comparison lists organizational controls that no individual plan — not even Pro+ or Max — includes: centralized license and policy management, audit logging, SAML SSO, and IP indemnification covering Copilot's suggestions. Enterprise goes one step further: it can index an organization's private codebase for more context-aware suggestions, adds a chat interface directly inside github.com, and is positioned to offer fine-tuned, private code-completion models for large customers, per GitHub's plan documentation.

If you're evaluating agent-heavy workflows at the team level — for instance, after reading about GitHub's agentic autofix and AI code scanning going GA — Business is the entry point where those org-wide controls start to matter; Free and even Pro/Pro+ are scoped to a single developer's account.

Who Should Actually Stay on Copilot Free?

Free makes the most sense if you fit one of these profiles:

• You code occasionally — side projects, learning, a few hours a week — and rarely approach 2,000 completions or 50 chat exchanges in a month.
• You mainly want inline autocomplete and only use chat sparingly for quick questions.
• You're evaluating whether Copilot fits your workflow before committing to a paid seat.
• You qualify for free Pro access instead — GitHub extends Pro, not just Free, at no cost to verified students, teachers, and maintainers of sufficiently popular open-source repositories, per GitHub's individual-plans documentation. That's a better deal than Free if you're eligible, since it removes the completion/chat caps entirely.

When — and to Which Plan — You Should Upgrade

The signal to upgrade is almost always hitting a wall, not a calendar date. Watch for these triggers:

• You're regularly capped on completions or chat before the month resets. That's the clearest sign Free's quota no longer matches your usage — move to Pro, which removes the completion cap entirely and replaces the 50-chat-request wall with a $15/month credit allowance that only chat, Edits and agent use draw from.
• You want to pick your model. If auto-selection is giving you inconsistent results and you want to deliberately choose a specific model for a hard problem, that requires Pro or above.
• You want the cloud coding agent — assigning an issue to Copilot and getting a pull request back without sitting at your keyboard — which is a Pro-and-up feature, not available on Free.
• You're burning through chat on premium models specifically (deep refactors, large context agent sessions). Pro+'s larger $70 credit pool and access to premium models like Claude's Opus-class tier, plus preview access to third-party coding agents, is built for that pattern.
• You run a team and need shared policy controls, SSO, audit logs or IP indemnity — that's when you move past individual plans entirely into Business or Enterprise, regardless of how much any one developer is using.

If you're also weighing Copilot against competing AI editors, our existing comparison of Copilot vs. Cursor pricing and agent capabilities covers how Copilot's paid tiers stack up against the alternative.

How to Check Your Current Usage Before You Hit the Wall

You don't have to guess when you're close to the Free plan's limits. GitHub's billing settings (under your account's Copilot settings on github.com) show a running count of completions and chat requests used in the current billing cycle. On paid plans, the same settings area shows AI Credits consumed against your base and flex allotments, with GitHub's documentation describing usage alerts at 75%, 90% and 100% of any configured budget — useful if you upgrade and want to avoid an unexpected overage charge once your included credits run out.

What to Do Next

If you've read this far because you're bumping into Free's ceiling, the practical next step is simple: open your GitHub billing settings, check how close you actually are to the 2,000-completion or 50-chat-request caps, and if you're regularly near either one, move to Pro — it's the smallest jump that removes the completion cap entirely and gives you a real, spendable credit allowance instead of a hard wall. If model choice, premium models, or a cloud agent matter more than raw volume, Pro+ or Max are worth pricing out directly. And if you're deciding for a team rather than yourself, skip the individual tiers altogether and start the conversation around Business or Enterprise, where the per-seat AI Credit pools and admin controls are designed to be managed centrally rather than per developer.

Written byFelix Moreau

Felix Moreau writes Pandromeda's software coverage and how-to guides. He covers Windows, macOS and Linux updates, the apps people rely on, emulators and developer tools, and turns official documentation into clear, numbered steps that work on the current version.

More from Felix MoreauAll Software & Guides coverage