OpenVLM MMT-Bench - Tech Engineering: leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-38B | 72.2 |
| 2 | GPT-4.1 (2025-04-14) | 61.1 |
| 3 | InternVL3-78B | 55.6 |
| 4 | Aquila-VL-2B | 55.6 |
| 5 | InternVL3-8B | 50 |
| 6 | Gemma 3 4B | 50 |
| 7 | GPT-4.1 Mini | 44.4 |
| 8 | Claude 3.7 Sonnet | 44.4 |
| 9 | Gemma 3 27B | 44.4 |
| 10 | Gemini 1.5 Pro | 44.4 |
| 11 | Gemma 3 12B | 44.4 |
| 12 | Qwen 2 VL 72B | 38.9 |
| 13 | InternVL2-8B | 33.3 |
| 14 | Qwen 2 VL 2B | 33.3 |
| 15 | GPT-4.1 Nano | 33.3 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-tech-engineering · How It Works · Data refreshed daily, snapshot 2026-09-05.