OpenAI unveils GPT-5.4 mini and nano for lower-cost, faster AI
OpenAI just rolled out its GPT-5.4 mini and nano models, aiming to deliver faster and smarter AI at a lower cost.
These new versions are built for quick responses (think coding help or real-time processing), and the mini model is now more than twice as fast as the last one, with better skills in coding, reasoning, and tool use.
GPT-5.4 mini hits 54.38% and 60%
The numbers back it up: GPT-5.4 mini scored noticeably higher than its predecessor on key benchmarks (54.38% on SWE-bench Pro and 60% on Terminal-Bench 2.0).
Pricing is pretty accessible too: the mini goes for $0.75 per million input tokens, while the nano is just $0.20 per million for simpler tasks via API.
As Aabhas Sharma, CTO at Hebbia, puts it, "GPT-5.4 mini delivers strong end-to-end performance for a model in this class. In our evaluations, it matched or exceeded competitive models on several output tasks and citation recall at a much lower cost. It also achieved higher end-to-end pass rates and stronger source attribution than the larger GPT-5.4 model."