ReactBench (Reaction Diagrams): leaderboard
ReactBench (2026) tests how multimodal models reason about the structure of chemical reaction diagrams drawn from journals and patents, with 1,618 expert-annotated questions scored by exact match. The score is the unweighted mean accuracy over four task dimensions, in percent, higher is better. It is unrelated to the ReactBench React UI coding benchmark.
Metric: Answer accuracy (%) on the four task dimensions (unweighted mean of localization, extraction, tracing and reasoning accuracy), over the 1,618 expert-annotated question-answer pairs on more than 1,300 real chemical reaction diagrams from chemistry journals and patents of ReactBench (single-line, multi-line, tree and graph topologies), multiple-choice or short open answers matched exactly against the key after template extraction, no LLM judge; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 24 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 32B Instruct | 73.91 | #606 |
| 2 | Gemini 3.1 Pro (Preview) | 73.88 | #99 |
| 3 | Qwen 3 VL 32B (Thinking) | 70.89 | #573 |
| 4 | GPT-5.5 | 69.59 | #54 |
| 5 | Qwen 2.5 VL 72B | 68.66 | #781 |
| 6 | Claude 3.5 Sonnet | 68.6 | #752 |
| 7 | Claude Sonnet 4.6 | 67 | #173 |
| 8 | Qwen-VL-Max (ReactBench checkpoint unspecified) | 65.03 | |
| 9 | GPT-4o | 63.83 | #751 |
| 10 | Qwen 3 VL 8B Instruct | 63.02 | #1029 |
| 11 | Gemini 1.5 Pro | 62.17 | #1063 |
| 12 | Qwen 3 VL 8B (Thinking) | 60.24 | #910 |
| 13 | Qwen 2.5 VL 7B | 57.32 | #1862 |
| 14 | InternVL2.5-78B | 53.74 | #1031 |
| 15 | InternVL2.5-8B-MPO | 50.38 | #2012 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=reactbench-reaction-diagrams · How It Works · Data refreshed daily, snapshot 2026-11-21.