I-WebGenBench - Chemistry: leaderboard
Metric: Overall rubric score (%) on the chemistry domain tasks; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 79.1 |
| 2 | Kimi K2.5 | 54.3 |
| 3 | Gemini 3.1 Pro (Preview) | 50.3 |
| 4 | MiniMax-M2.7 | 29.4 |
| 5 | Qwen 3 Max | 21.1 |
| 6 | DeepSeek V4 Flash | 17.2 |
| 7 | MiMo-V2.5 | 12.3 |
| 8 | DeepSeek V3.2 | 9.5 |
Interactive version: theaggregate.ai/benchmark?slug=i-webgenbench-chemistry · How It Works · Data refreshed daily, snapshot 2026-09-29.