Overview
GPT-5.5 Pro is a higher-compute version of GPT-5.5 designed for harder questions and work requiring greater accuracy. OpenAI describes it as using the same underlying model with parallel test-time compute. (openai.com)
Capabilities and uses
- Suitable for complex research, coding, data analysis and other multi-step professional tasks. It supports text and image inputs and produces text outputs. (openai.com)
- The API supports function calling, structured outputs, web search, file search, code interpreter, hosted shell and MCP. Image generation is available as a tool, rather than as a direct model output modality. (developers.openai.com)
- The documented context window is 1,050,000 tokens, with up to 128,000 output tokens. The knowledge cutoff is December 1, 2025. (developers.openai.com)
Availability and limitations
- Announced April 23, 2026; OpenAI's April 24 update states that GPT-5.5 Pro became available in the API. The documented default snapshot is
gpt-5.5-pro-2026-04-23. (openai.com) - API access uses the Responses API, including through Batch. Chat Completions, streaming, fine-tuning, audio and video are not supported. Some requests can take several minutes; OpenAI recommends background mode to avoid timeouts. Results should still be checked for accuracy. (developers.openai.com)
- At announcement, the model was rolling out in ChatGPT to Pro, Business and Enterprise users. This describes launch availability, not a guarantee of current plan access. (openai.com)
API pricing
As listed in OpenAI's model documentation, text-token pricing is USD $30 per 1 million input tokens and USD $180 per 1 million output tokens. There is no cached-input discount. Regional processing endpoints for data residency incur a 10% uplift. Batch is supported, but a separate GPT-5.5 Pro Batch rate was not verified. Tool usage may incur additional charges. Pricing source checked November 16, 2026. (developers.openai.com)