What GPT-5.5 is
On April 23, 2026, OpenAI officially announced GPT-5.5.
The company positions it as a model focused on:
- multi-step coding
- agentic work
- computer use
- knowledge work such as research, spreadsheets, and documents
From a product perspective, GPT-5.5 is the next major step after GPT-5.4, with a clear emphasis on long-horizon task execution and stronger autonomy.
Official source:
Where OpenAI says GPT-5.5 is stronger
According to OpenAI’s announcement, GPT-5.5 is especially aimed at:
- writing and debugging code
- researching information
- analyzing data
- creating documents and spreadsheets
- using tools to finish multi-step tasks
OpenAI also says GPT-5.5 keeps real-world serving latency close to GPT-5.4 while improving quality, especially in coding and knowledge work.
Is GPT-5.5 already in ChatGPT and the API
According to OpenAI:
- on April 23, 2026, GPT-5.5 started rolling out to
Plus,Pro,Business, andEnterpriseusers in ChatGPT and Codex - an April 24, 2026 update says
GPT-5.5andGPT-5.5 Probecame available in the API
So as of June 16, 2026, GPT-5.5 is no longer a limited preview. It is available across ChatGPT and the API based on OpenAI’s official announcement.
Who should care most about GPT-5.5
GPT-5.5 is especially relevant for:
- developers who want stronger coding performance
- teams building agent workflows
- people doing research or document-heavy work
- businesses that want AI to handle longer, messier tasks with less prompt micromanagement
For casual chat use, the difference may matter less than it does for technical or workflow-heavy use cases.
Conclusion
If you want the short version of OpenAI’s latest major model as of June 16, 2026, it is:
- GPT-5.5
- announced on April 23, 2026
- added to the API on April 24, 2026
This release is clearly positioned around practical AI work, especially coding, tool use, and multi-step task completion.