OpenAI: GPT-5.5 Pro

GPT-5.5 Pro is a text model from OpenAI, listed on api14 under the model id openai/gpt-5.5-pro. It sits in OpenAI's GPT-5.5 family as the high-capability variant, built for deep reasoning and accuracy on complex, high-stakes workloads. The provider release date is 2026-04-24, and its knowledge cutoff is 2025-12-01.

The model accepts text, images and files as input, and returns text only. Its context window is 1,050,000 tokens, with a maximum output of 128,000 tokens, so you can supply very large inputs and still leave room for a long response. Reasoning behaviour is exposed through request parameters rather than fixed, which matters if you need to trade depth against latency on a per-call basis.

Those capabilities point to workloads where the input is large and the answer has to hold up under scrutiny. Long document sets, mixed text and image material, and file attachments can go into a single request, which suits contract and policy review, technical analysis across many sources, research synthesis, and multi-step problem solving where the model has to reason through intermediate steps before committing to an answer. Structured outputs and tool use make it a reasonable fit for agent-style pipelines that need to call functions and return machine-readable results rather than free text.

You call it through api14's OpenAI-compatible API using the model id openai/gpt-5.5-pro, so existing OpenAI SDKs and clients work with a change of base URL and key. The supported request parameters are include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, seed, structured_outputs, tool_choice, tools and verbosity. In practice that means you can cap output length with max_tokens, control how much reasoning the model does with reasoning_effort (and inspect it with include_reasoning), constrain the reply shape with response_format or structured_outputs, and give the model callable functions through tools and tool_choice. The seed parameter allows you to request reproducible sampling where the backend supports it.

The main constraint to plan around is that output is text only, so anything requiring generated images, audio or video needs a different model. The 128,000 token output ceiling also means very long generations have to be split or continued across calls. If your task is short, latency-sensitive and does not benefit from deep reasoning or a million-token context, a smaller and faster text model on api14 will usually be the better choice.