GPAgentBench-2K - Missed Referral Rate: leaderboard
Metric: Missed referral rate (%; referral-required cases the agent managed locally, a safety failure; lower is better; zero-shot GP agent with six clinical actions over 2,428 physician-validated primary-care cases, GPT-4o-mini patient simulator, at most 10 turns). Source: arxiv.org. Saturation forecast: Around December 2027. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Kimi K2.6 | 40.3 |
| 2 | Gemini 3.1 Pro (Preview) | 47.8 |
| 3 | DeepSeek V4 Flash | 49.3 |
| 4 | GPT-5.4 | 51.2 |
| 5 | GLM-5.1 | 51.2 |
| 6 | DeepSeek V4 Pro | 57.2 |
| 7 | MiMo-V2.5-Pro | 59.2 |
| 8 | Lingshu-32B | 62.2 |
| 9 | Qwen 3.7 Max | 64.7 |
| 10 | MedGemma-27B | 65.7 |
| 11 | Claude Opus 4.7 | 67.2 |
| 12 | Claude Sonnet 4.6 | 77.1 |
| 13 | MiniMax-M2.7 | 78.1 |
| 14 | Lingshu-7B | 90 |
Interactive version: theaggregate.ai/benchmark?slug=gpagentbench-2k-missed-referral-rate · How It Works · Data refreshed daily, snapshot 2026-09-26.