GPT-6.1 Sol

What changed

GPT-6.1 Sol is the model listed in OpenAI’s September 29 release feed. The current model documentation describes it as offering near-Astra performance at a lower cost for coding, computer use and professional work. That is the provider’s positioning, not an independent result from AITrending.

Price and context

The documentation checked September 30 lists standard rates of $2 per million input tokens and $10 per million output tokens. Cached reads cost $0.10 and cache writes $2.50 per million tokens. These base rates do not apply unchanged to every request: prompts above 272,000 input tokens have higher rates, Fast mode costs twice Standard, and tool calls can have separate fees.

The listed context window is 1,050,000 tokens with up to 128,000 output tokens. Text and image inputs are supported; output is text. The listed knowledge cutoff is April 30, 2026.

Integration details

Use gpt-6.1-sol as the model ID. Tool calling uses the Responses API; Chat Completions support does not include tool calling. Reasoning levels run from low through max, with medium as the default. none and minimal are not supported.

Compare task completion, latency and total billed tokens against your existing model. A lower token price alone does not establish a lower cost per successful task.

Sources

• Official model documentation
• Dated release feed