GPT-5.5 Pro

GPT-5.5 Pro is a large language model developed by OpenAI as part of the GPT-5 family. The "Pro" designation places it at the higher-effort end of the family's lineup, above the standard and lighter-weight GPT-5 variants that OpenAI typically ships alongside it. Like other Pro-tier releases from OpenAI, it is aimed at requests where answer quality matters more than response latency. The model is multimodal on input, handling both text and images in the same request, which allows workloads such as reading screenshots, diagrams, charts, scanned documents, and UI mockups without a separate vision pipeline. Our catalog does not currently track a confirmed context window figure, per-token throughput, or time-to-first-token measurement for GPT-5.5 Pro, so those numbers should be checked against OpenAI's own model documentation before you size a workload around them. Where a provider on this page exposes the model, the exact limits and any reasoning-effort controls are set by that provider's API surface. In practice, Pro-tier models in the GPT-5 line are used selectively: routed the hardest subset of a workload — ambiguous debugging, multi-step analysis, long document synthesis — while cheaper siblings handle the bulk of traffic. Because Pro tiers generally spend more compute per request than mid-tier models, cost and latency per call tend to be higher, which makes a tiered routing setup the usual deployment pattern rather than using it for every request. Compare the providers listed in the pricing table on this page to see who currently serves it and at what rate.