GPT-5.5 Pro — Specs, Pricing & Limits

GPT-5.5 Pro's price is the second thing worth noticing about this model. The first is its entry-tier requests-per-minute ceiling, which sits at roughly a tenth of what every standard-tier current model gets — not a small discount off the usual number, a genuinely different order of magnitude.

That throughput ceiling is easy to miss if you're comparing models purely on price, since nothing about a high per-token rate signals a correspondingly tight request-rate limit. For any workload planning to call this model frequently rather than occasionally, the entry-tier RPM ceiling is worth checking before the price is, because it's the more likely thing to actually constrain a real deployment at low usage tiers.

No Fast mode is published for this model — consistent with the pattern across the Pro tier generally, where the premium price doesn't come with a matching latency option. And the long-context repricing rule that applies to several of its siblings isn't stated on this model's own page; that's recorded honestly as not published here, checked directly rather than assumed to carry over from the rest of its generation.

The table below is computed directly from this site's facts module: pricing across every published service tier and the rate-limit table — worth reading the RPM column specifically before assuming this model's constraints are purely about cost.