OpenAI releases "Spud" GPT-5.5 model
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Brendan Lynch/Axios
OpenAI on Thursday released its most capable model, GPT-5.5, codenamed "Spud," just one week after competitor Anthropic launched its latest model.
Why it matters: AI releases are getting faster, more efficient and more powerful.
What they're saying: "This is a new class of intelligence. It's a big step towards more agentic and intuitive computing," OpenAI co-founder Greg Brockman told reporters in a briefing.
- GPT-5.5 is a "faster, sharper thinker for fewer tokens" compared to 5.4, he added, and can handle multi-step workflows more autonomously with less user input.
- Despite the jump in capability, OpenAI says GPT-5.5 matches GPT-5.4's response speed in real-world use.
- It's available starting Thursday in ChatGPT and Codex for paid subscribers, with API access coming soon once OpenAI finishes incorporating additional cybersecurity guardrails.
Zoom in: OpenAI says the model improvements are strongest in coding, computer use, general office work and early scientific research, areas that require reasoning across longer contexts and executing tasks over time.
- Instead of step-by-step prompting, users can hand GPT-5.5 messy, multi-part tasks and let it plan, use tools, check its work and work toward a result.
- Teams with early access were able to gut check vibe-coded work, review thousands of additional documents and save up to 10 hours on work per week, according to the company.
The model was trained on Nvidia's GPUs, as with past OpenAI models.
- Nvidia employees will get access after a subset tested the model for several weeks.
- The model can act as a "chief of staff," helping power agents that are already acting as employees at Nvidia, vice president of enterprise computing Justin Boitano told Axios.
Zoom out: "We are moving to a compute-powered economy," Brockman added, referring to the idea that work will be powered by AI capacity, and therefore compute will become the bedrock of the economy.
- Nvidia says its new chips slash the cost of running advanced AI like GPT-5.5 up to 35x per token.
- That's an important metric for enterprises looking to increase AI usage without blowing out their IT budgets.
What we're watching: Whether this release leads to more enterprise adoption, a key focus for OpenAI.
- This comes after OpenAI executives called Anthropic's rise a "code red" "wake-up call" that resulted in a pivot in strategy to focusing on business customer adoption.
- Nvidia worked with OpenAI to create "a blueprint to make it easier for ... every enterprise to roll this out," Boitano said.
The bottom line: As AI systems take on longer, more complex tasks, the economics of running them may matter as much as the models themselves.
