OpenAI unveils GPT‑5.5
- OpenAI announced GPT‑5.5 on April 23, saying it improves coding, computer use, and deep research tasks. (cnbc.com) (axios.com) - Public details list a 1‑million‑token context window and “Thinking” and “Pro” variants, scoring 82.7% on Terminal‑Bench. (digitalapplied.com) - Coverage notes accelerating release cadence and contrasts OpenAI’s cybersecurity posture with rivals. (nytimes.com)
OpenAI on April 23 began rolling out GPT‑5.5, a new flagship model the company said is better at coding, using software, and handling long research tasks. (openai.com) Large language models predict the next word in a sequence, and newer versions are tuned to hold more context, use tools, and complete multi-step work with less step-by-step prompting. OpenAI said GPT‑5.5 is built for coding, online research, data analysis, documents, spreadsheets, and computer use across tools. (openai.com) OpenAI said GPT‑5.5 is rolling out in ChatGPT and Codex to Plus, Pro, Business, and Enterprise users, while GPT‑5.5 Pro is rolling out to Pro, Business, and Enterprise tiers. The company said the model matches GPT‑5.4 on per-token latency while using fewer tokens on the same Codex tasks. (openai.com) OpenAI’s published benchmarks put GPT‑5.5 at 82.7% on Terminal‑Bench 2.0, 78.7% on OSWorld‑Verified, and 84.4% on BrowseComp. OpenAI also listed a 1‑million‑token context window, which is the amount of text, code, or data a model can keep in view at once. (openai.com 1) (openai.com 2) The release came one week after Anthropic introduced Claude Opus 4.7, and Axios said OpenAI’s internal codename for GPT‑5.5 was “Spud.” The New York Times said the back-to-back launches show how quickly major model makers are now shipping new systems. (axios.com) (nytimes.com) CNBC reported that OpenAI is pitching GPT‑5.5 as stronger on “deeper research capabilities,” a category that has become central as chatbots move from question-answering to carrying out longer assignments. TechCrunch said the launch also fits OpenAI’s broader push to make ChatGPT and its tools act more like a software platform for work. (cnbc.com) (techcrunch.com) OpenAI paired the launch with a GPT‑5.5 system card and said it used internal and external red-teamers, added testing for advanced cybersecurity and biology capabilities, and gathered feedback from nearly 200 early-access partners. The company said GPT‑5.5 ships with its “strongest set of safeguards to date.” (openai.com 1) (openai.com 2) The New York Times contrasted that approach with some rivals that have moved faster to release powerful models while disclosing less about safety work. OpenAI’s pitch with GPT‑5.5 is that it can raise capability without slowing response times or loosening controls. (nytimes.com) (openai.com) The immediate test is whether users notice the gains outside benchmark charts. OpenAI has already started that test by putting GPT‑5.5 in front of paying ChatGPT customers one day after the April 23 announcement. (openai.com)
