LabOSBench - Preparation & Sample Handling: leaderboard

Metric: Subtask Success Rate (%; mean over the category's operation-level subtasks, 2 runs each). Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.

Top models

#ModelScore
1Seed-1.697.1
2uitars-1.5-7B91.2
3GPT-5.582.4
4Claude Opus 4.582.4
5Claude Sonnet 4.576.5
6Kimi K2.576.5

Interactive version: theaggregate.ai/benchmark?slug=labosbench-preparation-sample-handling · How It Works · Data refreshed daily, snapshot 2026-09-25.