OpenAI says GPT-6.1 Sol nears Astra at a fifth of the price

OpenAI on Tuesday released GPT-6.1 Sol, an upgrade to its GPT-6 Sol model that the company says nearly matches its top model, GPT-6 Astra, on agentic coding, computer use and professional work at one-fifth of Astra’s standard input and output token prices.

Who can use it. OpenAI says GPT-6.1 Sol is available starting today to all Plus, Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex. It “is not yet available in Chat,” the company says. Developers can reach it through the OpenAI API as gpt-6.1-sol.

What it costs. OpenAI lists standard API prices of $2 per million input tokens, $0.10 per million cached input tokens and $10 per million output tokens. The $2 and $10 rates are the same as GPT-6 Sol’s when it launched Sept. 22. Only the cached-input rate is lower: OpenAI says $0.10 is “50% less than GPT-6 Sol’s cached input pricing.” The company also says it will offer a faster version, GPT-6.1 Sol Ultrafast, “in the coming days,” with up to 8 times faster token generation than its standard speed in Codex.

What OpenAI claims. In benchmark results it published, OpenAI says the model matches Astra on the DeepSWE coding benchmark at roughly one-fifth of the cost. It says the model scores higher than Anthropic’s Opus 5.5 with fallbacks on the GDP.pdf document test, beats Opus 5.5 by 2.2 percentage points at medium reasoning effort on the AutomationBench workflow test, and comes within 2.1 points of Astra at maximum effort on the OSWorld computer-use test. On deliberately difficult prompts drawn from conversations where users had flagged an earlier model’s factual error, OpenAI says the share of answers with an error fell from 11.4% to 7.7% at low reasoning effort; the company says those prompts are not typical of everyday use. OpenAI says Astra still scores highest on its science test and should be used for the most difficult scientific research.

The risk rating. “Under our Preparedness Framework, we are treating GPT-6.1 Sol as Critical in cybersecurity and High for Biological and Chemical capability,” OpenAI says on its deployment safety page for the model. It says the model uses the same safeguards as Astra.

— Rowan Keel