# OpenAI topped web-dev Code generation and text-to-Image rankings

*Science. 63% certain. Thursday, 1 October 2026*

> Speculative fiction assembled from prediction-market prices. This has not happened.

**OpenAI also earned Second place in the AI Agent Lab and flagged a model as Critical, but lost the GPT trademark.**

OpenAI classified one of its models as having a Critical capability under its Preparedness Framework on Thursday. The same day the company was ranked the best provider of text-to-Image AI at the end of September. It also topped the Code Arena WebDev AI model rankings for the same period. In the AI Agent Lab competition OpenAI placed Second, and a U.S. trademark application for the term "GPT" was denied.

The Critical label signaled that the model possessed capabilities deemed especially powerful or risky. OpenAI applied the label as part of its internal Framework designed to assess and manage advanced AI functions.

Independent assessments placed OpenAI's text-to-Image system at the top of the leaderboard for September. The ranking reflected performance metrics such as Image fidelity, prompt adherence, and generation speed. No other company matched OpenAI's scores in the final week of the month.

In the Code Arena competition focused on web-development assistance, OpenAI's model secured the highest rank by the end of September. Judges evaluated the models on Code correctness, relevance to web-dev tasks, and integration ease. Competing offerings from Anthropic, SpaceXAI, and Alibaba trailed behind OpenAI's result.

The AI Agent Lab contest listed OpenAI as the Second-best participant, behind the leading entry. The ranking considered the agents' ability to plan, execute, and adapt to complex tasks. On the same day the U.S. Patent and trademark Office rejected OpenAI's request to register "GPT" as a trademark. The denial left the term unprotected despite the company's extensive use of the brand.

## Analysis

Classifying a model as Critical under the Preparedness Framework highlights OpenAI's internal risk-management approach.

Securing Second place in the AI Agent Lab demonstrates that OpenAI remains a strong contender in autonomous Agent performance, even if it fell short of the overall winner. The result suggests competitive pressure from other firms, indicating that future Agent benchmarks may see tighter margins and drive rapid innovation across the field.

The rejection of a GPT trademark leaves the widely recognized brand name without exclusive legal protection in the United States.

## What this was built from

- **Which company has the best Code Arena WebDev AI model end of September?**: OpenAI, 84%; settles Thursday, 1 October 2026 (polymarket). [Market](https://polymarket.com/event/which-company-has-the-best-code-arena-webdev-ai-model-end-of-september-20260717140116512)
- **Second-best AI Agent Lab end of September?**: OpenAI, 63%; settles Thursday, 1 October 2026 (polymarket). [Market](https://polymarket.com/event/second-best-ai-agent-lab-end-of-september-20260812152609626)
- **Which company has the best Text-to-Image AI end of September?**: OpenAI, 93%; settles Thursday, 1 October 2026 (polymarket). [Market](https://polymarket.com/event/which-company-has-the-best-text-to-image-ai-end-of-september-20260717133742993)
- **Will OpenAI classify any model as "Critical" capability under its Preparedness Framework before Oct 1, 2026?**: Yes, 96%; settles Thursday, 1 October 2026 (manifold). [Market](https://manifold.markets/Terminator2/will-openai-classify-any-model-as-c)
- **Will OpenAI be granted a trademark on "GPT" (in the U.S.)?**: No, 89%; settles Thursday, 1 October 2026 (manifold). [Market](https://manifold.markets/ScottLawrence/will-openai-be-granted-a-trademark)