ReactHuman - Semantic Action Accuracy: leaderboard
Metric: SAA (%). Source: arxiv.org. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 63.7 |
| 2 | Gemma 3 27B | 62.4 |
| 3 | Qwen 3 VL 235B A22B | 56.4 |
| 4 | Claude Opus 4.8 | 50.7 |
| 5 | Gemini 2.5 Flash | 48.7 |
| 6 | Kimi K2.6 | 40.9 |
Interactive version: theaggregate.ai/benchmark?slug=reacthuman-semantic-action-accuracy · How It Works · Data refreshed daily, snapshot 2026-09-20.