GitHub Copilot guide
Model costs, task choices, and plans. All in tables you can search and compare.
Reviewed against GitHub's documentation on . Prices are in USD.
What does each model cost?
GitHub pricing documentationCompare the token rates behind Copilot's AI credits. 1 AI credit = $0.01 USD. A task can make several model calls, so its cost depends on the tokens used across the whole session.
These are Copilot usage rates. Default and long-context tiers are separate; select a tier to see the corresponding input threshold. Paid code completions and next edit suggestions do not consume AI credits.
Scroll the table to compare all columns. Model names stay in view.
| Model / provider | Tier / input threshold | Input | Cached input | Cache write | Output | Availability |
|---|---|---|---|---|---|---|
| Claude Haiku 5.5 Anthropic | Default ≤ 100K input tokens | $0.10 | $0.01 | $0.125 | $0.50 | All paid plans |
| GPT-6 Luna OpenAI | Default ≤ 272K input tokens | $0.10 | $0.01 | $0.125 | $0.50 | All paid plans |
| GPT-6 Luna OpenAI | Long context > 272K input tokens | $0.20 | $0.02 | $0.25 | $0.75 | All paid plans |
| GPT-5.6 Luna OpenAI | Default ≤ 200K input tokens | $0.20 | $0.02 | $0.25 | $1.20 | All paid plans |
| MAI-Code-1.1-Flash Microsoft | Default | $0.20 | $0.02 | N/A | $1.20 | All paid plans |
| GPT-5.4 nano OpenAI · See note | Default | $0.20 | $0.02 | N/A | $1.25 | Codex extension |
| GPT-5.6 Luna OpenAI | Long context > 200K input tokens | $0.40 | $0.04 | $0.50 | $1.80 | All paid plans |
| GPT-5 mini OpenAI | Default | $0.25 | $0.025 | N/A | $2.00 | All paid plans |
| Claude Haiku 5.5 Anthropic | Long context > 100K input tokens | $0.50 | $0.05 | $0.625 | $2.50 | All paid plans |
| Gemini 3.7 Flash Google · Promo | Default | $0.75 | $0.075 | N/A | $3.75 | All paid plans |
| Gemini 3.8 Flash Google · Promo | Default | $0.75 | $0.075 | N/A | $3.75 | All paid plans |
| GPT-5.4 mini OpenAI | Default | $0.75 | $0.075 | N/A | $4.50 | All paid plans |
| Claude Haiku 4.5 Anthropic | Default | $1.00 | $0.10 | $1.25 | $5.00 | All paid plans |
| Grok 4.5 xAI | Default ≤ 200K input tokens | $2.00 | $0.50 | N/A | $6.00 | All paid plans |
| Grok 4.6 xAI | Default ≤ 200K input tokens | $2.00 | $0.50 | N/A | $6.00 | All paid plans |
| Grok 4.7 xAI | Default ≤ 200K input tokens | $2.00 | $0.50 | N/A | $6.00 | All paid plans |
| Claude Sonnet 5 Anthropic | Default | $2.00 | $0.20 | $2.50 | $10.00 | All paid plans |
| Claude Sonnet 5.5 Anthropic | Default | $2.00 | $0.10 | $2.50 | $10.00 | All paid plans |
| Gemini 4 Argon Google · Promo | Default | $2.00 | $0.10 | N/A | $10.00 | Premium plans |
| GPT-6 Sol OpenAI | Default ≤ 272K input tokens | $2.00 | $0.20 | $2.50 | $10.00 | Premium plans |
| GPT-6.1 Sol OpenAI | Default ≤ 272K input tokens | $2.00 | $0.10 | $2.50 | $10.00 | Premium plans |
| GPT-5.6 Terra OpenAI | Default ≤ 272K input tokens | $2.00 | $0.20 | $2.50 | $12.00 | All paid plans |
| Grok 4.5 xAI | Long context > 200K input tokens | $4.00 | $1.00 | N/A | $12.00 | All paid plans |
| Grok 4.6 xAI | Long context > 200K input tokens | $4.00 | $1.00 | N/A | $12.00 | All paid plans |
| Grok 4.7 xAI | Long context > 200K input tokens | $4.00 | $1.00 | N/A | $12.00 | All paid plans |
| GPT-5.3-Codex OpenAI | Default | $1.75 | $0.175 | N/A | $14.00 | All paid plans |
| Claude Sonnet 4.6 Anthropic · See note | Default | $3.00 | $0.30 | $3.75 | $15.00 | Annual Pro / Pro+ |
| GPT-5.4 OpenAI | Default ≤ 272K input tokens | $2.50 | $0.25 | N/A | $15.00 | All paid plans |
| GPT-6 Sol OpenAI | Long context > 272K input tokens | $4.00 | $0.40 | $5.00 | $15.00 | Premium plans |
| GPT-6.1 Sol OpenAI | Long context > 272K input tokens | $4.00 | $0.20 | $5.00 | $15.00 | Premium plans |
| Kimi K3 Moonshot AI | Default | $3.00 | $0.30 | N/A | $15.00 | All paid plans |
| GPT-5.6 Terra OpenAI | Long context > 272K input tokens | $4.00 | $0.40 | $5.00 | $18.00 | All paid plans |
| Claude Opus 5.5 Anthropic | Default | $4.00 | $0.20 | $5.00 | $20.00 | Premium plans |
| GPT-5.6 Sol OpenAI | Default ≤ 272K input tokens | $4.00 | $0.40 | $5.00 | $20.00 | Premium plans |
| GPT-5.4 OpenAI | Long context > 272K input tokens | $5.00 | $0.50 | N/A | $22.50 | All paid plans |
| Claude Opus 4.8 Anthropic | Default | $5.00 | $0.50 | $6.25 | $25.00 | Premium plans |
| Claude Opus 5 Anthropic | Default | $5.00 | $0.50 | $6.25 | $25.00 | Premium plans |
| GPT-5.5 OpenAI | Default ≤ 272K input tokens | $5.00 | $0.50 | N/A | $30.00 | Premium plans |
| GPT-5.6 Sol OpenAI | Long context > 272K input tokens | $8.00 | $0.80 | $10.00 | $30.00 | Premium plans |
| GPT-5.5 OpenAI | Long context > 272K input tokens | $10.00 | $1.00 | N/A | $45.00 | Premium plans |
| Claude Fable 5 Anthropic · See note | Default | $10.00 | $1.00 | $12.50 | $50.00 | Premium plans |
| Claude Fable 5.1 Anthropic · See note | Default | $10.00 | $0.25 | $12.50 | $50.00 | Premium plans |
| Claude Opus 4.8 (fast mode) (preview) Anthropic | Default | $10.00 | $1.00 | $12.50 | $50.00 | Premium plans |
| GPT-6 Astra OpenAI | Default ≤ 272K input tokens | $10.00 | $1.00 | $12.50 | $50.00 | Premium plans |
| GPT-6 Astra OpenAI | Long context > 272K input tokens | $20.00 | $2.00 | $25.00 | $75.00 | Premium plans |
No matching models. Try another name or provider, change the pricing tier, or reset the filters.
Premium plans means Pro+, Max, Business, and Enterprise. Organization policies and your client can further restrict access. N/A means GitHub lists no separate cache-write rate.
- Gemini 3.7 Flash and 3.8 Flash: promotional rates through 31 December 2026.
- Gemini 4 Argon: promotional rates until 1 December 2026.
- Claude Fable 5 and 5.1: Anthropic retains prompts and outputs for safety checks by default. Eligible enterprises can request a temporary zero-retention exemption through the end of 2026; see the model hosting source.
- Claude Sonnet 4.6: retired for monthly plans; remains available on annual Pro and Pro+ subscriptions.
- GPT-5.4 nano: not in Copilot Chat. GitHub specifies the Codex VS Code extension with Pro+; check current eligibility.
Rates: GitHub's pricing tables. Eligibility: supported models and plan restrictions. No separate long-context price is inferred where GitHub does not publish one.
Choose for the task
GitHub model guidanceStart with the least expensive model that handles your task reliably, then move up when the work needs more reasoning or autonomy. The matrix below is editorial guidance, not a benchmark or an official GitHub ranking.
- Fast
- GPT-6 Luna, Claude Haiku 5.5, Gemini 3.8 Flash
- Everyday
- GPT-5.6 Terra, Claude Sonnet 5.5
- Deep reasoning
- GPT-6.1 Sol, Claude Opus 5
- Long-running agent
- GPT-6 Astra, Claude Fable 5.1, Claude Opus 5.5
| Your task | Fast | Everyday | Deep reasoning | Long-running agent |
|---|---|---|---|---|
| Syntax questions and short explanations | Start here | Also fits | Not first choice | Not first choice |
| Boilerplate and repetitive edits | Start here | Also fits | Not first choice | Not first choice |
| Documentation and routine tests | Also fits | Start here | Not first choice | Not first choice |
| A feature with a clear, local scope | Also fits | Start here | Also fits | Not first choice |
| Debugging across several files | Not first choice | Also fits | Start here | Also fits |
| Architecture and difficult trade-offs | Not first choice | Also fits | Start here | Also fits |
| A coordinated codebase refactor | Not first choice | Also fits | Also fits | Start here |
| A long task with planning, tools, and verification | Not first choice | Also fits | Also fits | Start here |
These groups describe the work, not a hard boundary between models. Fable models have a data-retention exception. Check availability and review the result, including tests and security-sensitive changes.
Personal or business plan?
GitHub plan documentationPersonal allowances belong to the subscriber. Business and Enterprise seats contribute credits to a shared pool at the billing entity. Compare the monthly paid plans below.
| Compare | ProIndividual | Pro+Individual | MaxIndividual | BusinessOrganization | EnterpriseOrganization |
|---|---|---|---|---|---|
| Monthly price (USD) | $10 | $39 | $100 | $19 / seat | $39 / seat |
| Base AI credits / month | 1,000 | 3,900 | 10,000 | 1,900 | 3,900 |
| Flex credits / month | 500 | 3,100 | 10,000 | Not applicable | Not applicable |
| Total included AI credits | 1,500 | 7,000 | 20,000 | 1,900 | 3,900 |
| Who the credits belong to | Personal allowance | Personal allowance | Personal allowance | Pooled at the billing entity | Pooled at the billing entity |
| Model access | Selected models | Premium models | Priority access to premium models | Premium models, subject to policy | Priority access to premium models, subject to policy |
| Management | Personal settings | Personal settings | Personal settings | Central policies and seat management | Enterprise controls; requires Enterprise Cloud |
| After included credits run out | Set an extra-usage budget or upgrade | Set an extra-usage budget or upgrade | Set an extra-usage budget | Continues if additional usage is allowed | Continues if additional usage is allowed |
| GitHub training on interaction data | May be used; opt-out available | May be used; opt-out available | May be used; opt-out available | Not used by GitHub for model training | Not used by GitHub for model training |
Credits reset monthly. Included credits do not roll over. Personal flex allotments can change; they are separate from base credits. With additional usage disabled, organization usage stops after the pool is exhausted.
Free, Student, and annual subscriptions. Free and verified Student plans also exist, with auto model selection. Existing annual Pro and Pro+ subscriptions on legacy request billing have different rules; the table above covers monthly usage-based plans.
Training and retention are different. GitHub does not train on Business or Enterprise customer data. Personal subscribers have a training opt-out. Provider retention rules, including Fable exceptions, still need a separate check.
Details: personal credits and extra usage, shared pools and budgets, and model hosting and data use.
Sources & updates
This is a manually maintained reference, last reviewed on . GitHub can change rates, promotions, model availability, and allowances. Check the linked documentation before making a purchase or setting a budget.
Back to all resources