OpenAI Launches GPT-5.4 mini/nano: Performance Nears Flagship, Pricing Up to 4x Higher

OpenAI recently launched GPT-5.4 mini and nano models, targeting Agent, programming, and multi-modal workflow scenarios. The mini's API pricing is $0.75 per million tokens for input and $4.50 per million for output—2.25 to 3 times that of the previous GPT-5 mini. Nano's pricing is $0.20 input and $1.25 output, up to 4 times higher than its predecessor.

In terms of performance, GPT-5.4 mini scored 54.4% on the SWE-Bench Pro programming benchmark, close to the flagship GPT-5.4's 57.7% and significantly higher than GPT-5 mini's 45.7%. Both models support a 400,000-token context window. The mini also features computer control, tool calling, and web search, with a running speed more than twice as fast as the previous generation.

ChatGPT free and Go users will get the Thinking (reasoning) function driven by GPT-5.4 mini for the first time. The mini is also available on GitHub Copilot and Codex, consuming only 30% of GPT-5.4's quota in Codex. Nano is only available via API, suitable for light tasks like classification and data extraction.