Dental-TriageBench - Fine-Grained Micro-F1: leaderboard

Dental-TriageBench (2026) is an expert-annotated benchmark for multimodal dental triage: each outpatient case pairs a chief complaint with a panoramic radiograph, and a model predicts the full set of referral labels zero-shot. This board is micro-F1 on a 0 to 100 scale over the 22 fine-grained triage labels; higher is better.

Metric: Micro-F1 (from 0-1, times 100) over the 22 fine-grained triage labels on the 246 Dental-TriageBench outpatient cases (chief complaint plus panoramic radiograph), zero-shot multi-label referral prediction; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 19 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Flash (Preview)45.9#152
2Gemini 3 Pro (Preview)44.7#119
3GPT-5.240.4#185
4Claude Sonnet 4.535#261
5Kimi K2.534.8#266
6Grok 432.9#319
7Qwen-VL-Max-Latest (Dental-TriageBench checkpoint unspecified)28.3
8Ministral-3-14B-Instruct-251224.7#1639
9MiMo-VL-7B-RL17.5#1155
10MedGemma-27B-IT17.3#1350
11Qwen 2.5 VL 7B Instruct14.8#1902
12Qwen 3 VL 8B Instruct12.7#1029
13Lingshu-7B11.5#1635
14HuatuoGPT-Vision-7B10.1#2162
15LLaVA-OneVision-Qwen2-7B9.5#3035

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=dental-triagebench-fine-grained-micro-f1 · How It Works · Data refreshed daily, snapshot 2026-11-21.