OpenAI releases GPT-6.1 Sol for agentic coding and computer use
The model costs $2 per million input tokens, $10 per million output tokens, and $0.10 per million cached input tokens.
- Achieves 75.2% on DeepSWE v1.1 at high reasoning effort at approximately 76% lower cost per task than GPT-6 Sol.
- Scores 71.4% on the OSWorld 2.0 offline benchmark compared to Astra's 73.5%, at roughly one-seventh the cost per task.
- Scores 31.7% on AutomationBench at medium reasoning effort, up 4.8 percentage points from GPT-6 Sol.
- Reduces factual error rates by about 32% compared to GPT-6 Sol on difficult prompts at low reasoning effort.
Developers building agentic coding workflows and computer-use tools get higher benchmark performance alongside lower context-caching costs.
