BenchBench-Protocol - Genomics, Transcriptomics, and Sequencing: leaderboard
Metric: Normalized rubric score (%; weighted rubric elements met by the response to real protocol-modification tasks in the genomics, transcriptomics, and sequencing domain (19 tasks), graded by GPT-5.6 Terra; mean of repeated attempts per task). Source: arxiv.org. Saturation forecast: Around March 2028. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 5 (Max) | 56.1 |
| 2 | Kimi K3 (Max) | 47 |
| 3 | GPT-5.6 Sol (Max) | 43.8 |
| 4 | Grok 4.5 (High) | 41.7 |
| 5 | GPT-5.6 Terra (Max) | 40.5 |
| 6 | Inkling (Max) | 38.8 |
| 7 | GPT-5.6 Luna (Max) | 38.7 |
| 8 | Gemini 3.6 Flash (High) | 34.2 |
| 9 | GLM-5.2 (Max) | 30.6 |
Interactive version: theaggregate.ai/benchmark?slug=benchbench-protocol-genomics-transcriptomics-and-sequencing · How It Works · Data refreshed daily, snapshot 2026-09-26.