OpenAI has unveiled its new AI model GPT-5.5 and started distributing it to paying ChatGPT users.

On April 23 local time, TechCrunch reported that OpenAI introduced GPT-5.5 as its smartest and most intuitive model to use.

The company said GPT-5.5 can reason faster and more precisely with fewer tokens than GPT-5.4, enabling businesses and consumers to use more frontier AI.

OpenAI also stressed that the new model is progress toward its AI super app vision. Executives have mentioned plans to bundle ChatGPT, Codex and an AI browser into an integrated service for enterprise customers.

GPT-5.5 is designed for broad use, from core enterprise tasks such as agentic coding and knowledge work to experimental applications such as mathematics and scientific research. OpenAI said it scored higher on multiple benchmarks than previous models and rival models. It cited Google Gemini 3.1 Pro and Anthropic's Claude Opus 4.5 as comparators.

OpenAI has also been increasing the pace of new model releases. It released a new model last month, and also announced models in December and November last year. OpenAI Chief Scientist Jakub Pachocki said there would be significant improvements even in the short term, and very large improvements in the medium term.

Potential use in cybersecurity was also mentioned. Asked whether it would gain capabilities similar to Mithos, a cybersecurity tool recently unveiled by Anthropic, Mia Glaize from OpenAI's technical team said GPT-5.5 would have a major impact on the company's approach to deploying models in the digital security field.

GPT-5.5 will be rolled out from April 23 to ChatGPT Plus, Pro, Business and Enterprise users, and GPT-5.5 Pro will be available to Pro, Business and Enterprise users.