MedBench v5 - DataAgent: leaderboard

A Medical Atomic Skills environment of MedBench v5 (2026): multi-turn clinical data interaction over MySQL, PostgreSQL, CSV and clinical text: natural-language-to-SQL, query execution, trend and anomaly analysis, scored 0-100 by the environment metrics. Higher is better.

Metric: Environment score (0-100) on DataAgent (multi-turn clinical data interaction over MySQL, PostgreSQL, CSV and clinical text: natural-language-to-SQL, query execution, trend and anomaly analysis; accuracy and F1); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked. Updated 30 Sep 2026.

Top models

#ModelScoreOverall rank
1Kimi K2.682.67#181
2GPT-5.581.33#52
3Seed 2.0 Pro80.67#214
4Qwen 3.7 Max (Preview)79.33#94

No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.

Interactive version: theaggregate.ai/benchmark?slug=medbench-v5-dataagent · How It Works · Data refreshed daily, snapshot 2026-11-16.