CompBioBench v1 - Author Baselines: leaderboard
Metric: Accuracy (%, 100 agentic computational-biology tasks, mean of author runs). Source: huggingface.co. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 83.3 |
| 2 | Gemini 3.1 Pro (Preview) | 82 |
| 3 | Claude Opus 4.6 | 80.3 |
| 4 | Claude Opus 4.7 | 78.3 |
| 5 | Claude Sonnet 4.6 | 70 |
| 6 | Gemini 3 Flash | 66 |
| 7 | Claude Haiku 4.5 | 34 |
Interactive version: theaggregate.ai/benchmark?slug=compbiobench-v1-author-baselines · How It Works · Data refreshed daily, snapshot 2026-09-19.