MolViBench (Agent Collaboration): leaderboard
Metric: Pass@1 (%) over all 358 tasks under Agent Collaboration: a Coder and a Tester instance of the same model exchange test plans and PASS or FAIL feedback for up to 3 rounds, RDKit molecular coding tasks answered by generated Python code, official APIs, temperature 0, at most 4,096 output tokens; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 4 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro | 36 |
| 2 | Claude Opus 4.6 | 34.1 |
| 3 | DeepSeek V3.2 (Thinking) | 34.1 |
| 4 | Claude Opus 4.6 (Thinking) | 34.1 |
Interactive version: theaggregate.ai/benchmark?slug=molvibench-agent-collaboration · How It Works · Data refreshed daily, snapshot 2026-10-07.