GPT-5.5, OpenAI’s most recent AI model, was unveiled on Thursday. The company describes it as its “smartest and most intuitive to use model” to date. According to Greg Brockman, president and cofounder of OpenAI, the algorithm has improved skills in a number of areas and moves the company closer to developing its “super software.
Brockman described the new model as a significant step “towards more agentic and intuitive computing” on a call with reporters.
Brockman stated, “This model is a real step forward towards the kind of computing that we expect in the future but it is one step, and we hope to see many in the future.Compared to something like 5.4, it is a quicker, more accurate thinker for fewer tokens.This implies that businesses and consumers now have access to more cutting-edge AI, which is a component of our objective.
Additionally, Brockman stated that the concept was a further step toward developing a “super app”a multifunctional, Swiss Army knife-like program—which he and cofounder Sam Altman had previously discussed launching.The cofounders hope to help business clients by merging ChatGPT, Codex, and the AI browser into a single, cohesive offering.Notably, Elon Musk, Altman’s competition and former coworker at OpenAI, has expressed his desire to transform X into its own “super app.”
Only last month did OpenAI disclose its final model; it had previously released models in December and November.According to corporate employees, the corporation will continue to produce new models at a rapid pace for the foreseeable future.Jakub Pachocki, head scientist of OpenAI, stated, “We see pretty big gains in the short term, really significant improvements in the long term.” “To be honest, I think the past two years have been quite slow.
OpenAI claims that GPT5.5 is intended to be helpful in a wide range of fields, including more experimental AI applications like science and mathematics as well as fundamental enterprise domains like knowledge work and agentic coding.Additionally, data demonstrating the model’s higher performance across a variety of benchmarks was released by the company on Thursday.According to OpenAI, 5.5 routinely scores higher than its earlier models as well as models from rival Google and Anthropic (such as Gemini 3.1 Pro and Claude Opus 4.5).
The rivalry between OpenAI and Anthropic is frequently discussed, and GPT5.5 gave the two businesses another chance to contrast themselves.During the press conference, a reporter inquired as to whether GPT 5.5 would contain features akin to those of Anthropic’s recently released cybersecurity tool, Mythos(A report of unauthorized access to the application has caused controversy for Mythos in recent days.)According to OpenAI’s technical staff member Mia Glaese, GPT-5.5 will have a big influence on how the business uses its models for digital defense. “We have developed a robust and long-standing plan for our approach to cyber, and we have improved a sustainable approach to rolling out models safely,” Glaese stated.
OpenAI’s chief research officer, Mark Chen, stated that GPT5.5 outperformed its predecessors in computer work navigation and that the model “shows meaningful gains on scientific and technical research workflows,” adding that the company believes it could truly “help expert scientists make progress.” Chen added that it might help with medication discovery, a field that has garnered more attention from the business in recent years.
OpenAI claims that GPT 5.5 will be publicly accessible on Thursday.According to the company, ChatGPT’s Plus, Pro, Business, and Enterprise users are receiving the model, and Pro, Business, and Enterprise users will receive 5.5 Pro.
Leave feedback about this