OpenAI · Flagship

GPT-6 Astra API pricing and specs

OpenAI's most capable API model, with the full Responses API toolset: hosted shell, apply patch, skills, tool search, web and file search, computer use and remote MCP.

Input

⁨$10⁩

Cached input

⁨$1⁩

Output

⁨$50⁩

US dollars per million tokens, standard tier, short-prompt rate

Cache writes cost $12.50 per million. A separate Ultrafast tier runs at $60 input and $300 output, Fast mode doubles the price, and EU or other regional processing adds 10%. Astra has no 'none' reasoning setting and ignores custom temperature.

Prompts over 272K tokens are billed at ⁨$20⁩ input, ⁨$2⁩ cached and ⁨$75⁩ output per million, and the higher rate applies to the whole request.

Spec sheet

API model string: gpt-6-astra
Provider: OpenAI
Context: 1.05M tokens
Max output: 128K tokens
Released: September 8, 2026
Batch API: 50% off, results within hours
Documented agent features: Tool callingStructured outputComputer useRemote MCPBuilt-in web searchAdjustable reasoning

What a task costs

Four reference workloads priced at this model's list rates, with caching applied after the first call. Your numbers will differ; the calculator lets you change every assumption.

Workload	Per task	Per 1,000 tasks
Support bot 4 calls, short tool lookups	⁨$0.280⁩	⁨$280⁩
Coding agent 30 calls, files and test output	⁨$9.89⁩	⁨$9,887⁩
Browser agent 25 calls, screenshots and page text	⁨$6.98⁩	⁨$6,983⁩
Research agent 15 calls, long search results	⁨$6.62⁩	⁨$6,615⁩