OpenVLM LLaVA-Bench - Detail: leaderboard
OpenCompass OpenVLM LLaVA-Bench score for detailed visual description prompts.
Metric: Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 234 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4.1 (2025-04-14) | 103.7 | #250 |
| 2 | Gemma 3 27B | 102.9 | #607 |
| 3 | Gemma 3 4B | 96.6 | #1089 |
| 4 | GPT-4.1 Mini | 95.5 | #355 |
| 5 | Claude 3.7 Sonnet | 94.2 | #249 |
| 6 | Gemma 3 12B | 93.5 | #671 |
| 7 | Grok 2 (1212) | 93.1 | #496 |
| 8 | Gemini 2.0 Flash | 93 | #342 |
| 9 | Gemini 1.5 Pro | 91 | #445 |
| 10 | InternVL3-78B | 90.2 | #352 |
| 11 | InternVL3-38B | 83.5 | #405 |
| 12 | MiniCPM-V-2.6 | 83 | #836 |
| 13 | Claude 3.5 Sonnet (20241022) | 82.8 | #291 |
| 14 | GPT-4.1 Nano | 81.1 | #731 |
| 15 | Qwen 2.5 VL 7B | 80 | #643 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-llava-bench-detail · How It Works · Data refreshed daily, snapshot 2026-10-20.