MHGraphBench - Two-hop Verification: leaderboard
Two-hop verification questions, whether a drug relates to a disease through an intermediate disease, in MHGraphBench (2026), a knowledge-graph-grounded mental-health benchmark built from a curated PrimeKG subgraph, with letter-only multiple-choice answers. Accuracy in percent; higher is better.
Metric: Accuracy (%) on the 1,200 two-hop verification items (about half yes), 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 2030. 15 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 62.08 | #754 |
| 2 | GPT-4.1 | 57.58 | #506 |
| 3 | Llama3-Med42-8B | 55 | #5086 |
| 4 | BioMistral-7B | 53.92 | #8342 |
| 5 | Qwen 2.5 32B Instruct | 50.5 | #1253 |
| 6 | Llama 3.1 8B Instruct | 50.42 | #4655 |
| 7 | GPT-5 Mini | 50.25 | #373 |
| 8 | GPT-5.2 Instant | 50.17 | #309 |
| 9 | GPT-5.1 Instant | 50.17 | #463 |
| 10 | Qwen 2.5 7B Instruct | 50.08 | #2732 |
| 11 | DeepSeek R1 Distill Qwen 32B | 49.75 | #1844 |
| 12 | DeepSeek-R1-Distill-Qwen-7B | 49.75 | #7976 |
| 13 | Llama3-OpenBioLLM-8B | 49.75 | #6375 |
| 14 | Mistral 7B Instruct (v0.3) | 49.33 | #6759 |
| 15 | Meditron-7B | 49.17 | #8747 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mhgraphbench-two-hop-verification · How It Works · Data refreshed daily, snapshot 2026-11-22.