OpenAI adds GPT-5.5 and GPT-5.5 Pro to its API, including a 1M-token context window

A 1 million token context window and image input are now on the table for developers using OpenAI’s latest models, as Hacker News reported. ^[1]

OpenAI has released GPT-5.5 to both the Chat Completions API and the Responses API, while also introducing GPT-5.5 Pro specifically for Responses API requests. ^[1] The company describes GPT-5.5 as a “new frontier model for complex professional work.” ^[1]

Alongside the expanded API availability, GPT-5.5 is listed with a wide set of capabilities: a 1M token context window, image input, structured outputs, function calling, prompt caching, Batch support, tool search, built-in computer use, a hosted shell, apply patch, Skills, MCP, and web search. ^[1]

The release follows a run of announcements on April 23, 2026, when OpenAI announced GPT-5.5 as part of its ongoing cadence of model updates. ^[1]

OpenAI also updated its Agents SDK, adding features that include running agents in controlled sandboxes and inspecting and customizing the open-source harness. ^[1] The update also adds controls for when memories are created and where they’re stored. ^[1]

Separately, OpenAI released GPT Image 2, which it calls a “state-of-the-art image generation model for image generation and editing.” ^[1] GPT Image 2 supports flexible image sizes, high-fidelity image inputs, token-based image pricing, and Batch API support with a 50% discount. ^[1]