Cookie-Bench - React Dynamic: leaderboard

Cookie-Bench (2026) tests LLMs on front-end web development over 1,000 queries in 11 domains, with an autonomous evaluator that operates each generated page and scores the recorded session for functionality and aesthetics against human-calibrated rubrics, on a 0 to 100 scale. This board covers the interactive-application queries in React mode, where the model edits a React project inside an agent scaffold with file, build and web tools; higher is better.

Metric: Cookie total score (0-100) from the autonomous Cookie evaluator, which drives the rendered page and judges the continuous execution video against human-calibrated functionality and aesthetics rubrics, on the dynamic (interactive-application) queries, React mode: the model works inside an agent scaffold with file, build, web and image tools and edits a React project through tool calls; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Claude Opus 4.782.9#85
2Claude Opus 4.679.4#113
3Kimi K2.678.9#192
4GPT-5.476.4#135
5MiMo-V2-Pro75.2#353
6Qwen 3.6 Plus74.8#238
7GLM-5.174.6#198
8DeepSeek V4 Pro72.1#196
9Qwen 3.6 27B66.3#351
10DeepSeek V4 Flash66.2#267
11Gemini 3.1 Pro (Preview)64.3#99
12Gemini 3 Flash63.5#175
13Qwen 3.6 35B A3B59.8#481

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=cookie-bench-react-dynamic · How It Works · Data refreshed daily, snapshot 2026-11-21.