ToxReason - Kidney Toxicity: leaderboard
ToxReason (2026) tests mechanistic chemical toxicity reasoning: given a query molecule with experimental initiating-event evidence from structurally similar compounds, the model predicts liver, heart and kidney toxicity and explains how molecular initiating events lead to the adverse outcome along an Adverse Outcome Pathway. F1 for the kidney toxicity label, in percent; higher is better.
Metric: F1 (%) for kidney toxicity predicted by the model on the ToxReason test set (query chemicals with curated human CTD toxicity labels and adverse outcome pathway context, eight retrieved activation and inhibition examples from similar compounds), zero-shot, temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around January 2028. 11 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O3 | 60 | #237 |
| 2 | DeepSeek R1 Distill Llama 70B | 59.1 | #1669 |
| 3 | O4 Mini | 58.5 | #368 |
| 4 | Llama 3.1 70B Instruct | 57.1 | #1571 |
| 5 | Qwen 3 4B Instruct | 56.9 | #1696 |
| 6 | GPT-5 | 56.4 | #178 |
| 7 | GPT-4o | 54.8 | #751 |
| 8 | Gemma 3 27B (IT) | 54.2 | #1324 |
| 9 | GPT-5.1 | 50.3 | #240 |
| 10 | Qwen 2.5 14B Instruct | 48.8 | #1775 |
| 11 | Llama 3.1 8B Instruct | 31.2 | #4600 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=toxreason-kidney-toxicity · How It Works · Data refreshed daily, snapshot 2026-11-21.