SurgCoT - Micro-Transition Localization: leaderboard

Locating frame-level boundaries between surgical micro-phases in surgical videos of SurgCoT (2026), multiple-choice questions over 2,841 videos from seven specialties, answered zero-shot with only the video and the main question. Main-question accuracy in percent; higher is better.

Metric: Micro-transition localization main-question accuracy (%) on SurgCoT surgical-video multiple-choice questions (2,841 videos across 7 specialties and 35 procedures), baseline setting with only the video and the main question, zero-shot, temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Claude Sonnet 4.579.49#261
2Qwen 2.5 VL 7B76.59#1862
3Qwen 3 VL 8B74.07#917
4MedGemma-27B-IT72.69#1350
5HuatuoGPT-Vision-7B72.53#2162
6LLaVA-Med-7B71.43#6296
7Gemini 2.5 Pro69.82#276
8GPT-569.14#178
9InternVL3.5-8B68.35#1604
10Lingshu-7B65.12#1635

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

Interactive version: theaggregate.ai/benchmark?slug=surgcot-micro-transition-localization · How It Works · Data refreshed daily, snapshot 2026-11-21.