OpenAI’s top model, meant for the hardest reasoning, coding, computer-use and research work where cost matters less. It has a 1.05M tokens context window. Through OpenAI’s API it costs $10.00 per million input tokens and $50.00 per million output tokens.
Among the tools we track, it’s available in Codex, Copilot, Zed and Amp. Prompts over 272K input tokens cost twice as much for input and 1.5 times for output, and cache writes cost $12.50 per million tokens. Codex CLI made it the default for configs that don’t name a model in early September 2026.
Related changes
- Amp replaces its Dial with a per-thread model selector · November 12, 2026
- Codex CLI 0.153.4 makes GPT-6 Astra the default when no model is set · September 4, 2026
Sources
- developers.openai.com: gpt 6 astra
- learn.chatgpt.com: models
- docs.github.com: supported models
- zed.dev: models
- ampcode.com: models
Facts last checked November 15, 2026.