OpenAI Launches GPT-5.4 mini and GPT-5.4 nano, Performance Nears Full Version
OpenAI has recently launched two lightweight models, GPT-5.4 mini and GPT-5.4 nano. GPT-5.4 mini achieved a score of 54.4% in the SWE-Bench Pro coding benchmark, which is 3.3 percentage points lower than the full version GPT-5.4. Its running speed is twice as fast as the previous generation, and it supports a 400,000 token ultra-long context. The input price is only $0.75 per million tokens, and the output price is one-third of the full version. GPT-5.4 nano is positioned for ultra-lightweight tasks, with a price only one-twelfth of the full version.
OpenAI has recently launched two lightweight models, GPT-5.4 mini and GPT-5.4 nano. GPT-5.4 mini achieved a score of 54.4% in the SWE-Bench Pro coding benchmark, which is 3.3 percentage points lower than the full version GPT-5.4. Its running speed is twice as fast as the previous generation, and it supports a 400,000 token ultra-long context. The input price is only $0.75 per million tokens, and the output price is one-third of the full version. GPT-5.4 nano is positioned for ultra-lightweight tasks, with a price only one-twelfth of the full version.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.