Feature Comparison
| Plan or concept | Included monthly AI credits or status | Tracking note |
|---|---|---|
| Copilot Pro | 1,500 AI credits | Official allowance; verify before billing decisions. |
| Copilot Pro+ | 7,000 AI credits | Official allowance; verify before billing decisions. |
| Copilot Business | 1,900 AI credits; promotional 3,000 through September 1, 2026 | Track org billing policy and budget settings. |
| Copilot Enterprise | 3,900 AI credits; promotional 7,000 through September 1, 2026 | Track org billing policy and budget settings. |
| AI credit value | 100 credits = 1 USD of model usage | Credits are derived from model token pricing. |
| Legacy premium requests | Request-count language for older Copilot usage surfaces | Do not mix with token-priced AI credit math without checking docs. |
Calculator
Use the interactive calculator below to estimate AI credits from plan allowance, model, expected requests, and average tokens. The model price table is based on GitHub's public Copilot model pricing docs and should be rechecked before procurement.
How The Estimate Works
The calculator multiplies requests by average input tokens, output tokens, and cached prompt tokens, applies the selected model's public price per one million tokens, then converts estimated USD model cost into AI credits at 100 credits per USD.
estimated_usd = input_million_tokens * input_price + output_million_tokens * output_price + cached_million_tokens * cached_input_price estimated_ai_credits = estimated_usd * 100
Premium Requests Versus AI Credits
GitHub Copilot has used premium request language for higher-capability model calls, while newer usage-based billing docs describe AI credits for advanced model usage. Treat premium request counts as a workflow tracker and AI credits as the cost estimator when official docs price the model by tokens.
- Use requests to estimate how often the team will call an advanced model.
- Use AI credits to estimate token-priced model consumption.
- Check whether your plan, organization, and feature surface still shows premium request language.
- Record the reset date and last checked date in team docs.
Model Cost Table Source
Model prices and included plan credits can change. The calculator should be treated as last-checked guidance, not a billing guarantee.
Team Tracking Checklist
Use this page as a lightweight tracker before moving a team from legacy request counting to AI credit budgets.
[ ] Confirm plan allowance and promotional credits. [ ] Confirm which models are enabled for the organization. [ ] Estimate monthly requests and average tokens per request. [ ] Set budgets or alerts in GitHub billing where available. [ ] Review usage after the first billing cycle.
Legacy premium requests estimate: 200 request-style prompts. AI credit usage depends on token volume and model price, so compare this estimate with GitHub billing usage before making policy decisions.
Last checked: June 28, 2026. Selected model prices per 1M tokens: input $0.25, cached input $0.025, output $2.
FAQ
When do GitHub AI credits reset?
Plan allowances are described as monthly in GitHub Copilot billing docs, but teams should verify the exact reset and billing cycle in their GitHub billing settings.
Are Copilot premium requests the same as AI credits?
No. Premium requests are request-count language used in Copilot plan surfaces, while AI credits are a usage-based billing unit tied to model token pricing for supported advanced models.
How many AI credits do Copilot plans include?
As last checked on June 28, 2026, public GitHub docs list 1,500 for Pro, 7,000 for Pro+, 1,900 for Business, and 3,900 for Enterprise, with temporary higher promotional allowances for Business and Enterprise through September 1, 2026.
Why does my actual usage differ from this calculator?
Real usage can differ because prompts, cached tokens, tool calls, model routing, feature surfaces, organization policy, and billing updates can change the final count.