OpenAI Launches GPT-5.4 mini and nano Small Models: Low Latency and Performance Close to Full Version

OpenAI recently launched its most powerful small models to date, GPT-5.4 mini and GPT-5.4 nano, designed for high-frequency, latency-sensitive tasks. In scenarios like code assistance and real-time image reasoning, fast-responding small models are often more advantageous than large ones.

GPT-5.4 mini outperforms GPT-5 mini in code writing and logical reasoning, with over 2x faster running speed. Its performance in multiple benchmark tests is close to the larger GPT-5.4 model. It is suitable for fast-iterating code workflows and can quickly parse computer interface screenshots to complete operational tasks.

GPT-5.4 nano is the smallest and most affordable version, suitable for tasks like text classification and data extraction. Both models are now available: GPT-5.4 mini is integrated into API, Codex, and ChatGPT; GPT-5.4 nano is only available via API.