OpenAI Releases GPT-6.1 Sol at One-Fifth Astra's Price

OpenAI released GPT-6.1 Sol on September 29, 2026, as part of its DevDay product push, positioning it as an agentic coding model with near-Astra performance at one-fifth the price of GPT-6 Astra and a 95% discount on cached input pricing. The model is available via API and through ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Edu customers, but not in regular ChatGPT Chat. An Ultrafast mode running at 300 tokens per second is currently available for Astra and described as coming soon to GPT-6.1 Sol.

OpenAI's own benchmark data shows GPT-6.1 Sol scores 2.2 percentage points above Opus 5.5 on AutomationBench at medium reasoning effort and roughly one-third the cost, and more than doubles GPT-6 Sol's score on Terminal-Bench Science at maximum effort at less than half the cost per task. GPT-6.1 Sol costs $5.47 per task on Terminal-Bench Science versus $23.21 for Opus 5.5 and $23.80 for Astra. Third-party analysis by Handy AI notes that at maximum effort, GPT-6.1 Sol scores 36.1% on AutomationBench, trailing Opus 5.5 at 42.5% and GPT-6 Astra at 41.4%, and that OpenAI's stated advantage over Opus 5.5 is a medium-effort comparison. On DeepSWE, GPT-6.1 Sol scores 75.2%, ahead of Astra's best at a fraction of the cost, and 71.4% on OSWorld 2.0, 2.1 points under Astra. The system card rates GPT-6.1 Sol as Critical for cyber risk, and the model fails to disclose broken search tools in 2.1% of cases, down from 4.9% for GPT-6 Sol.