M3-VQA (Gold Entity Names): leaderboard

M3-VQA (CASIA, 2026) is a knowledge-based visual question answering benchmark of multimodal, multi-entity, multi-hop questions that require recognizing several entities and chaining evidence from Wikipedia, scored by the overlap between predicted and gold answer sets. This board is IoU accuracy in percent with the names of the entities behind the gold evidence added to the input; higher is better.

Metric: IoU accuracy (%) between the predicted and gold answer sets (exact match against Wikidata answer aliases, one-year margin for dates), averaged over the M3-VQA multi-entity, multi-hop knowledge-based visual questions, with the names of the entities behind the gold evidence of every hop added to the input; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-4o51.21#751
2Qwen 2.5 VL 72B Instruct45.05#841
3InternVL2.5-78B45.02#1031
4InternVL2.5-38B44.35#1373
5Qwen 2.5 VL 32B Instruct40.39#1062
6InternVL2.5-26B34.09#2156
7MiniCPM-V-2.633.6#3267
8Qwen 2.5 VL 7B Instruct31.83#1902
9LLaVA-OneVision-7B31.69#2981
10DeepSeek-VL230.56#2927
11Qwen 2 VL 7B Instruct30#2573
12InternVL2.5-4B29.56#2549
13InternVL2.5-8B25.71#2370
14Qwen 2.5 VL 3B Instruct24.98#2713
15InternVL2.5-2B23.2#6468

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

Interactive version: theaggregate.ai/benchmark?slug=m3-vqa-gold-entity-names · How It Works · Data refreshed daily, snapshot 2026-11-21.