Introducing GPT-5.4 mini and nano

OpenAI has announced the release of GPT-5.4 mini and nano, its newest small language models designed for speed and efficiency in high-volume applications. GPT-5.4 mini demonstrates significant improvements over GPT-5 mini in areas like coding, reasoning, and multimodal understanding, achieving performance levels approaching the larger GPT-5.4 model. GPT-5.4 nano, the most compact and cost-effective version, is ideal for tasks prioritizing speed and affordability, such as classification and data extraction. These models are particularly well-suited for applications requiring low latency, like coding assistants and computer-using systems. Availability includes the API, Codex, and ChatGPT, with varying pricing based on usage. Customers testing the models have reported strong performance in coding workflows, especially for fast iteration and targeted edits.

(Source：OpenAI)