Benchmarks.bio - MetagenomicsBench: leaderboard
MetagenomicsBench (benchmarks.bio) — agentic microbial-community analysis over real metagenomic data (shotgun, 16S/amplicon, long-read, metatranscriptomic and spatial assays): community structure, host associations, function and metabolism, longitudinal dynamics and interventions across 100 evaluations with three runs each, scored as pass rate with deterministic grading.
Metric: Pass Rate (%). Source: benchmarks.bio. Saturation forecast: Around February 2027. 22 models tracked. Updated 19 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 | 62.7 | #5 |
| 2 | Claude Fable 5.1 | 61.7 | #15 |
| 3 | Claude Sonnet 5.5 | 60 | #19 |
| 4 | Claude Opus 5 | 54 | #20 |
| 5 | Claude Opus 4.8 | 50 | #58 |
| 6 | GPT-6 Astra | 49.7 | #6 |
| 7 | Grok 4.7 | 48.7 | #47 |
| 8 | Claude Sonnet 5 | 42.3 | #111 |
| 9 | GPT-6 Sol | 42 | #26 |
| 10 | Grok 4.6 | 42 | #45 |
| 11 | DeepSeek V4.1 Flash | 38 | #62 |
| 12 | Claude Opus 4.7 | 35.3 | #78 |
| 13 | Gemini 3.7 Flash | 31.7 | #34 |
| 14 | GPT-5.6 Sol | 31.3 | #30 |
| 15 | GPT-6 Luna | 31 | #99 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #10 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=benchmarks-bio-metagenomicsbench · How It Works · Data refreshed daily, snapshot 2026-11-16.