ToxReason - Toxicity Prediction: 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. Macro F1 over the three organ toxicity labels, in percent; higher is better.

Metric: Macro F1 (%) over liver, heart and kidney toxicity labels 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 2029. 11 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1DeepSeek R1 Distill Llama 70B65.7#1676
2GPT-564.7#185
3O363.8#242
4Llama 3.1 70B Instruct63.5#1579
5GPT-4o63.2#754
6O4 Mini61.8#372
7Qwen 3 4B Instruct61.8#1711
8GPT-5.160.1#248
9Gemma 3 27B (IT)60#1341
10Qwen 2.5 14B Instruct57.4#1788
11Llama 3.1 8B Instruct49#4655

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=toxreason-toxicity-prediction · How It Works · Data refreshed daily, snapshot 2026-11-22.