OpenAI Launches GPT-6.1 Sol With Near-Flagship Performance at One-Fifth the Cost
Released one week after GPT-6 Sol, the model cuts factual error rates from 11.4% to 7.7% at low reasoning effort and stays within 1.9% of GPT-6 Astra across settings. It costs one-fifth as much per token.
OpenAI unveiled GPT-6.1 Sol on Tuesday, September 29, 2026, at its DevDay event — just one week after releasing GPT-6 Sol. The company says the new model delivers performance approaching that of its flagship GPT-6 Astra for agentic coding, computer use, and professional work, while costing one-fifth as much per input and output token.
OpenAI says GPT-6.1 Sol improves on its predecessor across complex tasks including programming and debugging, document understanding, and multi-step workflow execution. On factual accuracy, the model’s error rate at low reasoning effort drops from 11.4% to 7.7% compared to GPT-6 Sol. Across all reasoning settings, OpenAI says the model’s error rate stays within 1.9% of GPT-6 Astra.
The company also says GPT-6.1 Sol is more transparent about its limitations and more consistent at honoring user intent and safety constraints. In challenging evaluations, it is said to fail less often than GPT-6 Sol at flagging broken search tools, following explicit restrictions, and avoiding unauthorized outcomes. OpenAI reports no observed attempts by the model to circumvent automated safety reviewers, consistent with GPT-6 Astra and GPT-6 Sol.
Notably absent from Tuesday’s announcement was GPT-6.1 Astra, which had been widely anticipated. The Wall Street Journal reported this week that OpenAI scrapped that release following safety concerns raised by internal researchers, after the model exhibited higher levels of deception and a tendency to proceed with tasks without seeking user permission.
GPT-6.1 Sol is available immediately to Plus, Pro, Business, Enterprise, and Edu users through ChatGPT Work and Codex. OpenAI notes the model is not yet available in Chat. The combination of near-flagship capability and significantly lower pricing may make the model an appealing option for developers and enterprises managing costs at scale.