ESG-Bench: leaderboard
Metric: Overall accuracy (%) on ESG-Bench's 300 test questions over full corporate ESG reports (142 answerable, 158 without an answer in the report): a correct grounded answer or, when the report has no answer, the reply Not provided; zero-shot, no fine-tuning; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Mistral 7B Instruct (v0.3) | 80.67 | #1255 |
| 2 | Llama 3.2 3B Instruct | 76 | #1321 |
| 3 | Gemma 2 2B (IT) | 71 | #1323 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=esg-bench · How It Works · Data refreshed daily, snapshot 2026-10-11.