GPT-5.4 Mini — Specs, Pricing & Limits

GPT-5.4 Mini is the mid-tier member of the GPT-5.4 generation — priced well under the full GPT-5.4 model, with a genuinely smaller context window rather than the same window at a discount.

That smaller window sits comfortably under the long-context repricing threshold that governs several larger models on this roster, and no repricing rule is published for this model at all — checked directly against its own page rather than assumed absent because the window looks small enough that it "probably doesn't apply." For any workload that fits inside this model's window in the first place, that's one less pricing mechanic to plan around compared to its larger siblings.

It shares its rate-limit group with the cheapest member of the current GPT-5.6 generation rather than with the rest of the GPT-5.4 family — worth checking directly rather than assuming rate limits track model generation the same way pricing does, since here they clearly don't.

Fast mode is published for this model, at a real premium over its already-low Standard rate, which makes it one of the more affordable ways to get lower latency on this site's entire roster even after the premium.

The table below is computed directly from this site's facts module: pricing across every published service tier and the rate-limit table.