OpenAI · Current flagship
GPT-6 Astra API pricing & cost
Current token rates, context limits and direct paths into TokenCOGS workload calculators for OpenAI’s current flagship API model.
Current rates
Pricing snapshot
Fallback snapshot verified September 16, 2026; client checks the TokenCOGS live registry on load.
Input / 1M tokens$10
Output / 1M tokens$50
Cached input / 1M$1
Context window1.05M tokens
Model facts
Context and quick cost example
What is tracked
- Released September 3, 2026.
- Maximum output: 128,000 tokens.
- Prompts above 272,000 input tokens use OpenAI’s published 2× input/cache and 1.5× output long-context pricing rule.
- Cache writes are tracked at 1.25× the standard input rate in the registry.
2K input + 500 output tokens
At current standard text-token rates, one request with 2,000 input and 500 output tokens is about $0.045 before tool charges or long-context premiums. 1,000 requests ≈ $45.00.
Use your workload