OpenAI’s Latest AI Models Are Built for Speed
The latest models for ChatGPT users and developers using OpenAI’s API are designed to be workhorses, designed for tasks like ambient coding, where large, powerful AI models are expensive. GPT-5.4 mini and nano, released by OpenAI on Tuesday, are the smallest and fastest versions of GPT-5.4, which debuted earlier this month.
These new models are part of an effort by OpenAI to lean into coding as it battles with rival Anthropic for the AI software engineering market. OpenAI’s latest models for its Codex coding software directly challenged Anthropic’s Claude Code, which went viral in late 2025 for its ability to create apps from scratch.
Disclosure: Ziff Davis, the parent company of CNET, filed a lawsuit in 2025 against OpenAI, alleging that it had violated Ziff Davis’ copyrights in the training and operation of its AI systems.
OpenAI claims that GPT-5.4 mini is more than twice as fast as its predecessor, GPT-5 mini, for tasks such as coding, reasoning, and using tools. On some benchmarks, it is close to the capabilities of the standard GPT-5.4 model.
Learn more: Best AI chatbots of 2026
The company suggested that GPT-5.4 mini is best for things like editing and debugging code. It could be used as a subagent in Codex, with a larger model like GPT-5.4 delegating some tasks to the faster and cheaper model.
GPT-5.4 nano is even smaller and OpenAI suggested it for tedious jobs like classification and data mining.
GPT-5.4 mini will be available to developers via the API and via Codex and ChatGPT. ChatGPT Free and Go users can access it via the “Reflection” feature. Others will find it as a fallback model when they reach the rate limit for GPT-5.4 Thinking. GPT-5.4 nano is only available in the API.
