OpenAI launches GPT-5.4 mini and nano for faster coding help
Technology
OpenAI just rolled out mini and nano versions of its latest AI model, GPT-5.4.
These smaller models are built for speed: think coding help or apps that need instant replies.
The mini packs almost the same punch as the full version but runs faster.
GPT-5.4 mini 54.38% on SWE-bench Pro
The GPT-5.4 mini scored big on benchmarks (54.38% on SWE-bench Pro) and is more than twice as fast as GPT-5 mini, costing $0.75 per million input tokens and $4.50 per million output tokens.
The nano is even lighter (great for simple stuff like sorting or extracting info) at just $0.20 per million input tokens and $1.25 per million output tokens, making these options way more affordable than the full GPT-5.4 model if you're watching your budget or building something quick.