ARC-Easy (accuracy) — AI2 ARC-Easy split, accuracy metric.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | DeepSeek-V3 | DeepSeek | 98.9 | 1 | 2026-05-01 |
| 2 | Llama 3.1 Instruct 405B | Meta | 98.4 | 1 | 2026-05-01 |
| 3 | Qwen2.5 Instruct 72B | Alibaba | 98.4 | 1 | 2026-05-01 |
| 4 | DeepSeek-V2 | DeepSeek | 97.6 | 1 | 2026-05-01 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.