OpenAI has released GPT‑5.5, a model the company describes as “our smartest and most intuitive to use model yet” and “the next step toward a new way of getting work done on a computer.”
The model has been built to take on tasks that span multiple steps and tools without requiring close oversight from the user. According to OpenAI, users can “give GPT‑5.5 a messy, multi-part task and trust it to plan, use tools, check its work, navigate through ambiguity, and keep going.”
The areas where OpenAI says gains are strongest include agentic coding, computer use, knowledge work, and what the company terms “early scientific research” – fields where, in OpenAI’s words, “progress depends on reasoning across context and taking action over time.”
OpenAI GPT‑5.5 uses fewer tokens and runs faster – Here’s how it works
GPT‑5.5 is capable of writing and debugging code, conducting research online, analysing data, creating documents and spreadsheets, operating software, and moving across tools to complete a task from start to finish.
One of the notable claims OpenAI makes about GPT‑5.5 concerns its performance relative to GPT‑5.4. The company states that GPT‑5.5 “matches GPT‑5.4 per-token latency in real-world serving” despite operating at “a much higher level of intelligence.”
It also “uses significantly fewer tokens to complete the same Codex tasks,” making it, in OpenAI’s words, “more efficient as well as more capable.”
OpenAI says GPT‑5.5 is being released with “our strongest set of safeguards to date, designed to reduce misuse while preserving access for beneficial work.”
The company evaluated the model across its full suite of safety and preparedness frameworks. It also worked with internal and external red-teamers, added targeted testing for what it describes as “advanced cybersecurity and biology capabilities,” and gathered feedback from “nearly 200 trusted early-access partners before release.”
GPT‑5.5 is now rolling out to Plus, Pro, Business, and Enterprise users in ChatGPT and Codex. GPT‑5.5 Pro is available to Pro, Business, and Enterprise users in ChatGPT.
API access is not yet live. OpenAI has acknowledged that “API deployments require different safeguards” and that the company is “working closely with partners and customers on the safety and security requirements for serving it at scale.” The company added that it will “bring GPT‑5.5 and GPT‑5.5 Pro to the API very soon.”
