OpenAI rolls out GPT-5.5
- OpenAI announced GPT-5.5, saying the retrained base model is stronger at coding, using computers, and deep research. - Early benchmarks showed an 82.7% score on an agentic coding test, and the model is rolling to paid users. - OpenAI also launched ChatGPT for Clinicians to reduce hospital paperwork by assisting with documentation and research. (cnbc.com) (indiatoday.in)
OpenAI has started rolling out GPT-5.5, a new model the company says is better at coding, web research, and carrying out multi-step work on a computer. (openai.com) OpenAI published the release on April 23, 2026 and said GPT-5.5 is available in ChatGPT and Codex for Plus, Pro, Business, and Enterprise users, while a GPT-5.5 Pro version is going to Pro, Business, and Enterprise accounts. (openai.com) The company said GPT-5.5 matched GPT-5.4 on response speed per token in real-world serving, while using fewer tokens on Codex tasks and posting higher scores on several internal and external tests. On Terminal-Bench 2.0, which measures agent-style coding work, OpenAI reported an 82.7% score for GPT-5.5 versus 75.1% for GPT-5.4. (openai.com) These systems are built to do more than answer a single prompt. OpenAI described GPT-5.5 as a model that can plan, use tools, check its own work, move across software, and keep working through a longer task instead of stopping after one reply. (openai.com) That pitch lines up with OpenAI’s recent product push toward “agentic” software, which means an artificial intelligence system can take actions across apps and websites on a user’s behalf. In the same week, OpenAI also announced workspace agents in ChatGPT and new developer tools for agent workflows. (openai.com) OpenAI paired the model launch with a healthcare product aimed at a narrower audience. On April 22, the company said ChatGPT for Clinicians would be free for verified U.S. physicians, nurse practitioners, physician assistants, and pharmacists. (openai.com) OpenAI said that version is designed for clinical documentation, evidence review, and medical research at the point of care, and its help documentation says it includes trusted clinical search, citations, reusable skills, and deep research across medical literature. (openai.com 1) (openai.com 2) The healthcare launch builds on OpenAI for Healthcare, a program the company introduced earlier in 2026 for hospitals and health systems that want enterprise controls and support for Health Insurance Portability and Accountability Act compliance. (openai.com) OpenAI said it tested GPT-5.5 under its safety and preparedness frameworks, added targeted checks for cybersecurity and biology risks, and gathered feedback from nearly 200 early-access partners before release. (openai.com) The rollout puts OpenAI’s latest model in front of paying ChatGPT users immediately, while its clinician product shows where the company is trying to turn those broader model gains into job-specific tools. (cnbc.com) (openai.com)
