Back to all plans
GitHub Copilot
Global standard
GitHub Copilot coding extension — the world's most widely used AI coding assistant, native to VS Code; usage-based AI Credits billing across all plans since June 2026
IDE ToolsGitHub Copilot
Core models
GPT-5 miniGPT-5.3-CodexGPT-5.4GPT-5.4 miniGPT-5.4 nanoGPT-5.5Claude Haiku 4.5Claude Sonnet 4.5Claude Sonnet 4.6Claude Opus 4.5Claude Opus 4.6Claude Opus 4.6 (fast mode)Claude Opus 4.7Claude Opus 4.8Gemini 2.5 ProGemini 3 FlashGemini 3.1 ProGemini 3.5 FlashMAI-Code-1-FlashRaptor mini
GPT-5.5
OpenAI's flagship and the most powerful GPT variant in Copilot—suited to complex multi-step reasoning, large-scale refactors, and long debug chains; uses long-context pricing beyond 272K input tokens.
GPT-5.4
Versatile flagship balancing speed and capability; handles the vast majority of everyday coding scenarios within 272K context—a strong cost-effective pick for chat and agent workflows.
GPT-5.4 mini
A faster, lighter variant of GPT-5.4 with lower Credits cost—recommended for quick Q&A, completion-linked interactions, and high-frequency small tasks where latency matters.
GPT-5.4 nano
The lightest GPT model in Copilot—minimal Credits per request, suited to high-volume low-complexity code comprehension and text generation so lightweight calls don't drain your premium-model budget.
GPT-5.3-Codex
The flagship Codex successor optimized for coding—strong on code generation, function comprehension, and API-call patterns; Copilot's go-to pick for intensive code-centric workflows.
See the official site for more models
Additional core model names still appear above, with full details on the latest official page.
Plan details
Usage
Includes 2,000 monthly code completions plus a limited allowance for chat, agent tasks, and premium-model exploration—ideal for first-time users who want to understand Copilot's core workflow before committing to a paid plan.
Models
Base GPT and Claude models are available on the free tier, though premium models like Opus or GPT-5.5 and sustained agent workflows are noticeably capped.
Highlights
For anyone wanting to feel Copilot's in-editor impact before subscribing, the free tier covers most introductory evaluation—the core code completion experience remains intact.
Best for
Individual developers trying an AI coding assistant for the first time
Usage
Unlimited completions plus 1,500 AI Credits per month (1,000 base + 500 flex), covering daily Q&A, code explanation, test generation, and light agent tasks—the most natural starting point for individual developers using Copilot as a primary tool.
Models
Freely switch among GPT-5.5, Claude Opus, Gemini 3.5 Flash, and all 20 supported models—well suited for developers who want to leverage different model strengths within their daily workflow.
Highlights
The flex allotment adjusts over time as AI economics evolve while the 1,000 base credits stay fixed; overages bill automatically at per-token rates with no interruption to your workflow.
The Student plan shares Pro-level benefits with unlimited code completions; eligible students can apply through GitHub Education for a free upgrade.
The Student plan shares Pro-level benefits with unlimited code completions; eligible students can apply through GitHub Education for a free upgrade.
Best for
Individual developers using Copilot as a primary daily tool
Usage
7,000 AI Credits per month (3,900 base + 3,100 flex)—nearly 5× Pro's total—enough to make premium models like Claude Opus 4.8 and GPT-5.5 part of real daily workflows rather than occasional experiments.
Models
Better suited to heavy individual developers running complex code comprehension, cross-file refactors, test generation, and extended debugging chains—a larger budget meaningfully reduces the hesitation to use premium models.
Highlights
If you often downgrade models or stop agent tasks on Pro because of credit limits, the difference at Pro+ is immediately practical—the extra capacity removes that friction.
Best for
Power individual developers who frequently use premium models and agent workflows
Usage
20,000 AI Credits per month (10,000 base + 10,000 flex) behaves more like a sustained personal AI development budget pool—agent workflows, code review, complex refactors, and extended explanations are all effectively on by default.
Models
For developers with multiple parallel projects, frequent branch experiments, and continuous reliance on premium model judgment, the Max capacity difference is very concrete—it may be overkill for average users but makes a clear difference for heavy users.
Highlights
At $100/month, the per-interaction cost of premium model calls becomes effectively invisible—the right choice for individuals who genuinely treat Copilot as a core workbench for all-day intensive work.
Best for
All-day power users running multiple projects in parallel
Usage
Per-seat billing with approximately 3,000 AI Credits/seat/month (base plus flex) pooled at the organization level for centralized budget management and member access control.
Models
Includes full model catalog access, cloud agent tasks, centralized license assignment, policy management, and usage auditing—the first true step from personal Copilot to formal team deployment with full org governance.
Highlights
Admins can set model policies and spending caps for the whole organization and track team-wide Credits consumption in real time via the usage dashboard.
Best for
Development teams needing centralized control over seats, budgets, and model access
Usage
Builds on Business with approximately 7,000 AI Credits per seat per month (base plus flex) pooled at the enterprise level—suited to large engineering organizations with high-frequency premium-model and agent usage.
Models
Deeper GitHub Enterprise Cloud integration covering org policies, compliance auditing, priority model access, and enterprise SAML/SCIM governance—the default choice for broad Copilot deployments in Enterprise Cloud environments.
Highlights
Once Copilot is treated as a formal productivity system rather than a personal plugin, the gap between Enterprise and Business in credit capacity, governance depth, and compliance becomes very clear.
Best for
Large engineering organizations requiring strong governance, priority capabilities, and enterprise compliance
Quota
Product type
IDE Tools
Official product line
GitHub Copilot
5-hour limit
—
Weekly limit
—
Monthly limit
—
Notes
- 1 AI Credit = $0.01 USD; base credits are fixed 1:1 to your subscription price, while the flex allotment is a variable add-on that adjusts over time; usage beyond the full allowance bills automatically at per-token model rates.
- Monthly Pro/Pro+ subscribers migrated automatically to usage-based billing on June 1, 2026; annual subscribers stay on PRU-based billing until expiry and then move to Free, or can convert to a monthly plan before expiry with prorated credit for the remaining annual value.
- Business ($19/seat) and Enterprise ($39/seat) pool their credits at the organization or enterprise level rather than isolating per individual; admins can set a budget cap and monitor usage in real time via the dashboard.
Supported coding tools
VS CodeVisual StudioJetBrainsXcodeEclipseGitHub CLI
Pricing and model data sourced from official vendor websites
FAQ
General7 条