OpenVLM MMT-Bench - Electronic Object Recognition Leaderboard — The Aggregate

OpenVLM MMT-Bench - Electronic Object Recognition: leaderboard

Metric: Score (%). Source: huggingface.co. 207 models tracked.

Top models

#	Model	Score
1	GPT-4.1 Mini	100
2	InternVL3-78B	100
3	InternVL3-8B	100
4	Qwen 2 VL 72B	100
5	Qwen 2 VL 7B	100
6	InternVL3-38B	100
7	Qwen 2 VL 2B	100
8	Pixtral-12B	100
9	Llama 3.2 11B Instruct	100
10	Grok 2 (1212)	100
11	Gemma 3 12B	100
12	GPT-4.1 (2025-04-14)	100
13	Llama 3.2 90B Vision Instruct	100
14	InternVL2-8B	95
15	Aquila-VL-2B	95

Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-electronic-object-recognition · How It Works · Data refreshed daily, snapshot 2026-09-05.