MedProbeBench - Evidence Verification: leaderboard
Metric: Evidence verification (x100): task-weighted composite of claim recall, search effectiveness (reference recall) and factual consistency of cited claims, for LLMs with search tools on MedProbeBench: each system writes a complete clinical guideline for 50 WHO-classification tumor entities (one run per task), compared with the expert guideline; higher is better. Source: arxiv.org. Saturation forecast: Around March 2028. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 | 43.7 |
| 2 | Grok 4 | 41.4 |
| 3 | Gemini 3 Flash (Preview) | 40.1 |
| 4 | GPT-5 | 37.4 |
| 5 | GPT-4.1 | 35.6 |
| 6 | Claude Sonnet 4 | 33.6 |
| 7 | Claude Sonnet 4 (Thinking) | 29.7 |
Interactive version: theaggregate.ai/benchmark?slug=medprobebench-evidence-verification · How It Works · Data refreshed daily, snapshot 2026-10-07.