OpenAI launches GPT-5.1 for coding and agents
GPT-5.1 launched November 12 with 400K context, SWE-bench Verified 76.3%, priced at $1.25/$10 per 1M tokens.
OpenAI launched GPT-5.1 on November 12, the first major update in the GPT-5 family, targeting coding and multi-step agentic tasks. The model keeps a 400K context window, priced at $1.25 input and $10 output per 1M tokens, with prompt caching cutting input cost by 90%.
GPT-5.1 scores 76.3% on SWE-bench Verified, with improved instruction following and math over GPT-5. It ships in five variants: Instant for general chat, Thinking with low/medium/high reasoning, Codex Mini and Codex Max for automated coding pipelines, and Pro replacing GPT-5 Pro for $200/month subscribers.
Codex variants added an apply_patch tool (editing code without JSON escaping) and shell execution, directly competing with agentic coding tools like Devin and Replit Agent. On November 19, OpenAI added GPT-5.1-Codex-Max and GPT-5.1 Pro variants.