OpenAI just released its newest model, GPT-5.5, and the company is not being shy about what it can do.
On Thursday, co-founder and president Greg Brockman called it the company’s “smartest and most intuitive” model to date.
That’s a bold claim, but OpenAI has the benchmark data to back it up. So what does GPT-5.5 actually do better? A lot.
It thinks faster, uses fewer tokens to get there, and opens doors for both everyday users and big enterprise clients.
This model is a building block. OpenAI is working toward something far more ambitious, a “super app” that could change how we use AI entirely.
Also read: OpenAI Is Building a Desktop Superapp
GPT-5.5
Brockman described the model as “a faster, sharper thinker for fewer tokens.” In plain terms? It gets to better answers more quickly, and it costs less compute to do it.
That matters a lot for businesses; fewer tokens means lower costs at scale. And for consumers, it means snappier responses in daily use.
OpenAI also says GPT-5.5 is notably better at navigating computers. Mark Chen, the company’s chief research officer, said the model shows real gains in “scientific and technical research workflows.” He even pointed to drug discovery as an area where it could make a meaningful difference.
Beyond science, the model targets agentic coding and deep knowledge work. These are the kinds of tasks that enterprises care about most.
These organizations regularly need help with automating complex workflows, writing and debugging code, and synthesizing large volumes of information.
Super App
Brockman used Thursday’s launch to signal OpenAI’s longer-term vision.
The company wants to build a “super app,” one unified platform that combines ChatGPT, Codex, and an AI browser.
Think of it like a Swiss Army knife for digital work. It will be one app and many tools, all powered by the same underlying AI.
Brockman said GPT-5.5 is “a real step forward towards the kind of computing that we expect in the future.” He added that more steps are coming, and soon.
The super app idea is not entirely new. Sam Altman has talked about it before, but it’s gaining momentum now.
Interestingly, Elon Musk, Altman’s former colleague at OpenAI, is chasing a similar vision with X.
OpenAI
GPT-5.5 dropped just weeks after OpenAI’s last model release. And that release came not long after the one before it.
The company has been churning out new models at a pace that surprises even industry insiders.
Jakub Pachocki, OpenAI’s chief scientist, put it bluntly during the press briefing.
“We see pretty significant improvements in the short term, extremely significant improvements in the medium term.
In fact, I would say the last two years have been surprisingly slow.” – Jakub Pachocki, Chief Scientist, OpenAI
Competition
OpenAI released benchmark data alongside the launch. According to those numbers, GPT-5.5 outperforms both Gemini 3.1 Pro from Google and Claude Opus 4.5 from Anthropic across a range of tests.
Of course, companies usually publish data that flatters their own products. Independent testing will tell a fuller story over the coming weeks.
Still, the comparison to Anthropic came up directly during the press call.
A journalist asked whether GPT-5.5 would have features similar to Mythos, Anthropic’s recently announced cybersecurity tool, which has faced scrutiny following a report of unauthorized access.
Mia Glaese, a technical staff member at OpenAI, said the company has a “strong and longstanding strategy” for deploying AI safely in cybersecurity contexts.
Access
GPT-5.5 is widely available as of Thursday. OpenAI is rolling it out to Plus, Pro, Business, and Enterprise users inside ChatGPT.
A more powerful version, GPT-5.5 Pro, is heading to Pro, Business, and Enterprise accounts. Free-tier users are not included in the initial rollout, at least not yet.
