Benchmarks.bio - VariantBench: leaderboard
VariantBench (benchmarks.bio) — agentic genetic-variant discovery and interpretation over real sequencing data (short- and long-read WGS and exome across Illumina, PacBio and ONT): small-variant, structural-variant and repeat calling with QC, statistical and population genomics, and clinical/personal genome interpretation, scored as pass rate with deterministic grading.
Metric: Pass Rate (%). Source: benchmarks.bio. Saturation forecast: Around June 2028. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5 | 49.72 | #12 |
| 2 | GPT-6 Astra | 48.6 | #6 |
| 3 | Grok 4.6 | 46 | #26 |
| 4 | Claude Opus 4.8 | 42.1 | #37 |
| 5 | GPT-5.6 Sol | 42.1 | #19 |
| 6 | GPT-5.6 Terra | 38.7 | #41 |
| 7 | Claude Sonnet 5 | 38.7 | #57 |
| 8 | Grok 4.5 | 38.1 | #48 |
| 9 | Claude Opus 4.7 | 37 | #49 |
| 10 | Gemini 3.5 Flash | 35.3 | #60 |
| 11 | Gemini 3.1 Pro (Preview) | 34.5 | #58 |
| 12 | Kimi K3 | 32.8 | #30 |
| 13 | GPT-5.5 | 30.2 | #31 |
| 14 | GPT-5.6 Luna | 29.9 | #68 |
| 15 | Grok 4.3 | 20.3 | #188 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=benchmarks-bio-variantbench · How It Works · Data refreshed daily, snapshot 2026-10-20.