OpenAI Releases GPT-5.5 for ChatGPT and Codex

TL;DR: OpenAI released GPT-5.5 on April 23, 2026, with availability beginning in ChatGPT and Codex. The model is aimed at real work that spans coding, online research, data analysis, documents, spreadsheets, and tool use. For teams comparing AI tools, the meaningful shift is not just a higher benchmark number; it is OpenAI pushing a model that can carry more of a multi-step workflow with less manual steering.

What changed with GPT-5.5?

GPT-5.5 is OpenAI's newest frontier model for complex work. In the launch post, OpenAI frames it as a model that understands tasks earlier, uses tools more effectively, checks work, and keeps going across longer workflows.

That matters because many AI products are moving from single-turn chat into agentic workflows: code changes, browser work, spreadsheet analysis, report drafting, and research synthesis. Readers comparing products in the AI tools directory should watch how quickly GPT-5.5 appears inside coding assistants, research tools, and enterprise workflow apps.

Who gets access first?

OpenAI says GPT-5.5 is rolling out to Plus, Pro, Business, and Enterprise users in ChatGPT and Codex. GPT-5.5 Pro is rolling out to Pro, Business, and Enterprise users in ChatGPT.

Codex users get GPT-5.5 with a 400K context window, and OpenAI also describes a Fast mode for Codex that generates tokens faster at a higher cost. This is most relevant for teams using Codex for larger repositories, longer-running tasks, or code review loops that need more context than a small prompt can carry.

Why this is more than a benchmark update

OpenAI reported gains across coding, professional work, computer use, academic tasks, and tool-use benchmarks. The headline for builders is that GPT-5.5 is being positioned as a work model, not only a chat model.

For AI software buyers, that changes the evaluation question. Instead of asking only "Which model answers best?", teams should ask whether the model can use tools reliably, explain its plan, handle files and context, and finish a task without constant re-prompting.

What to watch next

The practical story now moves to adoption. If GPT-5.5 improves coding and tool use in the real world, expect changes across AI coding assistants, data analysis products, research agents, and enterprise automation tools. We will track those changes in AI News and compare adoption signals on the AI leaderboard.