Benchmarks.bio - TxBench-PP: leaderboard
Benchmarks.bio TxBench-PP evaluates agentic perturbation/transcriptomics analysis workflows with deterministic grading over realistic biological data-analysis tasks.
Metric: Pass Rate (%). Source: benchmarks.bio. Saturation forecast: Around January 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.8 | 59.33 | #37 |
| 2 | GPT-5.5 | 55.33 | #31 |
| 3 | Gemini 3.5 Flash | 51.33 | #60 |
| 4 | GPT-5.4 | 49.67 | #79 |
| 5 | Claude Opus 4.7 | 49.33 | #49 |
| 6 | Claude Opus 4.6 | 44.67 | #63 |
| 7 | Gemini 3.1 Pro (Preview) | 40 | #58 |
| 8 | Claude Sonnet 4.6 | 36 | #90 |
| 9 | Kimi K2.6 | 29.67 | #104 |
| 10 | Grok 4.20 0309 (Reasoning) | 19.67 | #112 (Grok 4.20 0309) |
| 11 | Grok 4.3 | 18.33 | #188 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=benchmarks-bio-txbench-pp · How It Works · Data refreshed daily, snapshot 2026-10-20.