AI BriefingIntroducing GPT-5.4 mini and nano
Key point
OpenAI has released GPT-5.4 mini and nano, high-performance small models optimized for coding and subagent tasks.
Details
OpenAI has released GPT-5.4 mini and nano, high-performance small models optimized for coding and subagents tasks. These models are designed for handling large-scale workloads, with a focus on maximizing speed and efficiency.
GPT-5.4 mini shows significantly improved coding, reasoning, multimodal understanding, and tool-use capabilities compared to the existing GPT-5 mini, and is more than 2x faster. In particular, it demonstrates outstanding efficiency on major benchmarks such as SWE-Bench Pro and OSWorld-Verified, approaching the performance of the higher-tier GPT-5.4 model.
The smallest and most affordable GPT-5.4 nano is designed for coding subagents that perform classification, data extraction, ranking, and simple auxiliary tasks.
These models are optimized for environments where low-latency is key, such as:
- Coding assistants that require fast responses
- Subagents that perform auxiliary tasks in parallel
- Computer-using systems that interpret screenshots
- Multimodal applications that require real-time image reasoning
GPT-5.4 mini is now available in the API, Codex, and ChatGPT. It supports a 400k context window in the API, priced at $0.75 per 1 million input tokens and $4.50 per 1 million output tokens.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.