Knowledge benchmark
LABBench2 leaderboard
LABBench2: An Improved Benchmark for AI Systems Performing Biology Research. Every model the catalog carries a published LABBench2 value for, ranked by that value.
A benchmark of realistic biology-research tasks involving literature, figures, tables, databases, and bioinformatics files.
LABBench2 ranking
6 models with a published LABBench2 value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Aggregate accuracy |
|---|---|---|---|
| 1 |  Gemini 3.8 Flash | 86.2 | |
| 2 |  Claude Opus 5 | Anthropic | 84.2 |
| 3 | Gemini 3.7 Flash | 82.1 | |
| 3 |  GPT-5.6 Sol | OpenAI | 82.1 |
| 5 | GPT-5.6 Terra | OpenAI | 81.2 |
| 6 | Claude Sonnet 5 | Anthropic | 80.1 |
Evidence key: Observed
Rows are ordered by the value LABBench2: An Improved Benchmark for AI Systems Performing Biology Research published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.