MHGraphBench - Two-hop Verification with Evidence: leaderboard
Metric: Accuracy (%) on the 1,200 two-hop verification items with sanitized PrimeKG feature snippets added, letter-only answers; the OpenAI API models answer at temperature 0 (at most 120 completion tokens) with strict answer-letter parsing, the open models are scored by forced-choice option-letter log-probabilities; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 | 71.42 |
| 2 | GPT-4o | 68.83 |
| 3 | Qwen 2.5 32B Instruct | 61.25 |
| 4 | GPT-5.2 Instant | 61.08 |
| 5 | GPT-5.1 Instant | 60 |
| 6 | GPT-5 Mini | 59.67 |
| 7 | Qwen 2.5 7B Instruct | 56.75 |
| 8 | Mistral 7B Instruct (v0.3) | 56.17 |
| 9 | Llama 3.1 8B Instruct | 50.25 |
| 10 | DeepSeek-R1-Distill-Qwen-7B | 49.75 |
| 11 | DeepSeek R1 Distill Qwen 32B | 49.58 |
Interactive version: theaggregate.ai/benchmark?slug=mhgraphbench-two-hop-verification-with-evidence · How It Works · Data refreshed daily, snapshot 2026-10-07.