SAIL-VL-2B: benchmark results
Provider: ByteDance. Released 2025-02-21. Access: Open.
Aggregate Rating 1424 ± 20, rank #4969 of 10916 rated models, from 518 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMT-Bench - Social Relation Recognition | 80 | Score (%) | 99.3 |
| OpenVLM MMT-Bench - Depth Estimation | 85 | Score (%) | 98.1 |
| OpenVLM SEEDBench IMG - Text Understanding | 79.8 | Accuracy (%) | 97.8 |
| OpenVLM MMT-Bench - Disaster Recognition | 100 | Score (%) | 96.8 |
| OpenVLM MMT-Bench - Geometrical Perspective | 91.7 | Score (%) | 96.6 |
| OpenVLM MMT-Bench - LVLM Response Judgement | 65 | Score (%) | 96.4 |
| OpenVLM MMT-Bench - Geometrical Relativity | 75 | Score (%) | 96.1 |
| OpenVLM MMT-Bench - Human Interaction Understanding | 90 | Score (%) | 95.9 |
| OpenVLM MMT-Bench - RR | 80 | Score (%) | 95.6 |
| OpenVLM MMT-Bench - HLN | 76.2 | Score (%) | 93.2 |
| OpenVLM MMT-Bench - Other Biological Attributes | 85 | Score (%) | 93.2 |
| OpenVLM AI2D - Volcano | 100 | Accuracy (%) | 92.8 |
Interactive version: theaggregate.ai/model?slug=sail-vl-2b · How It Works · Data refreshed daily, snapshot 2026-11-08.