What is GPT-6.1 Sol?
GPT-6.1 Sol is an OpenAI model for complex coding, computer use and professional work, which OpenAI describes as near-Astra performance at a lower cost. It sits between GPT-6.1 Terra and GPT-6 Astra on price. Use it for implementing features, multi-step agent work and research where Terra is not enough but Astra's price is hard to justify.
What can you do with it?
OpenAI lists GPT-6.1 Sol for complex coding, computer use and professional work. In practice that means feature work across several files, agents that operate applications, and research or reports that need more reasoning than GPT-6.1 Terra offers.
OpenAI released GPT-6.1 Sol on September 29, 2026, describing it in the API changelog as built for complex coding and professional work at a lower cost than GPT-6 Astra. The same changelog notes that Sol supports Multi-agent in beta, which lets the model delegate parts of a Responses API request to subagents.
Sol, Terra or Astra?
Within OpenAI's current lineup, Sol costs twice as much per token as GPT-6.1 Terra and one fifth of GPT-6 Astra's $10 input and $50 output. In ChatGPT Work and Codex, OpenAI's Plus estimates are 15 to 160 local messages per five-hour window for Sol against 5 to 45 for Astra, so the same allowance stretches roughly three times further.
OpenAI's own guidance on the Terra launch says Sol and Astra remain its recommendation for the most complex coding tasks. Choose Terra for high-volume screen and document work, Sol for most hard engineering, and Astra when Sol at high effort still fails your checks. OpenAI also says Astra at Low effort can outperform Sol at High, so compare usage rather than effort labels.
Try it on a small task
Suggested workflow: pick one feature ticket in a repository you can reset with git. In Codex or the Responses API, run gpt-6.1-sol at the default medium effort and ask it to plan first, then implement and run the tests.
Check that the diff stays within the files the ticket names and that tests pass. If it misses requirements, clarify the ticket before raising effort; OpenAI notes that higher effort cannot supply missing information. Compare usage with the same ticket on GPT-6.1 Terra if cost matters.
Pricing and access
In the API, the model ID is gpt-6.1-sol. Standard pricing is $2 per million input tokens, $0.10 for cached input, $2.50 for cache writes and $10 per million output tokens. Prompts over 272K input tokens cost twice the input and cache rates and 1.5 times output for the full request. Batch and Flex are 50% lower, Fast mode is 2x and Ultrafast 6x Standard, and regional processing adds 10%. Knowledge cutoff is April 30, 2026.
In ChatGPT, GPT-6.1 Sol is a Work and Codex model, selectable in the model picker where your account has it. OpenAI's usage guide estimates 15 to 160 local messages per five-hour window on Plus and Standard Business. Regular Chat uses GPT-6 Sol for its thinking levels, a different model. Ultrafast in ChatGPT is limited at launch to Pro $500 and eligible Enterprise and Edu workspaces.
What stands out
- 1.05M-token context window, 922K max input and 128K max output
- Reasoning effort low, medium (default), high, xhigh and max
- Text and image input; no audio or video
- Ultrafast mode in the Responses API at 6x Standard prices
Before you choose
- Twice the token price of GPT-6.1 Terra.
- Not the model behind regular ChatGPT chats.
- No audio or video input.
Go to the source
Sources & further reading
Based on the official documentation below. Features and access can change; check the provider before making a decision.
Was this page useful?
Help us make our coverage more useful. This is feedback on the page, not a rating of the tool.
Go a little further