GPT-5.5 "Spud" Is Here: OpenAI's Most Autonomous Model Yet Takes Aim at Anthropic's Enterprise Crown

Six weeks. That's all it took OpenAI to go from GPT-5.4 to GPT-5.5, codenamed "Spud." The new model is rolling out today to Plus, Pro, Business, and Enterprise users in ChatGPT and Codex.

GPT-5.5: built to work with less hand-holding

The headline feature? Autonomy. You can hand GPT-5.5 a messy, multi-part task and walk away. The model plans, picks its tools, checks its own work, and pushes through ambiguity until the job is done. Greg Brockman called it "a big step towards more agentic and intuitive computing."

Performance gains cluster in four areas: agentic coding, computer use, knowledge work, and early scientific research. On Terminal-Bench 2.0, GPT-5.5 scores 82.7%, a new state-of-the-art among publicly available models.

GPT-5.5 vs Claude Opus 4.7 and Gemini 3.1 Pro

According to VentureBeat, GPT-5.5 leads on 14 benchmarks, compared to 4 for Claude Opus 4.7 and 2 for Google's Gemini 3.1 Pro. But the picture is more nuanced than that: on Humanity's Last Exam without tools, GPT-5.5 Pro scores 43.1%, behind Opus 4.7 at 46.9%. Raw reasoning is still contested territory.

The enterprise angle is what really matters here. OpenAI has reportedly been in an internal "Code Red" state since December 2025, watching Anthropic's B2B revenue surge. GPT-5.5, with its focus on agentic workflows and office productivity, is OpenAI's answer.

Smarter, not slower, but pricier per token

A key selling point: despite the capability jump, GPT-5.5 matches GPT-5.4's per-token latency. OpenAI claims the model uses fewer tokens to complete equivalent tasks, reducing total cost. Per-token pricing is higher, though, so developers will need to do the math.

On safety, OpenAI tested GPT-5.5 with nearly 200 trusted early-access partners. The model is classified "High" risk for cyber and bio capabilities under OpenAI's Preparedness Framework, and API access is delayed while additional guardrails are put in place.

The bottom line on GPT-5.5

ChatGPT now claims 900 million weekly active users and 50 million paying subscribers. Those are staggering numbers. GPT-5.5 may not settle the "who's best" debate once and for all, but it proves OpenAI won't cede the enterprise crown to Anthropic without a serious fight.