Phi-3.5-Vision: benchmark results
Provider: Microsoft. Released 2024-08-20. Access: Open.
Aggregate Rating 1426 ± 14, rank #4843 of 10916 rated models, from 1054 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MEGA-Bench Task - Bridge Strategies Advanced | 57.9 | Task Score (%) | 97.7 |
| MEGA-Bench Task - MFC Bench Check Clip Stable Diffusion Generate | 64.3 | Task Score (%) | 96.2 |
| OpenVLM MMT-Bench - Traffic Anomaly Detection | 60 | Score (%) | 94.9 |
| MEGA-Bench Task - Position Relationship | 80 | Task Score (%) | 94.2 |
| MEGA-Bench Task - Red Teaming Celebrity | 89.5 | Task Score (%) | 94.2 |
| OpenVLM MMT-Bench - 3D CAD Recognition | 75 | Score (%) | 94.2 |
| MEGA-Bench Task - Graph Hamiltonian Path | 44.6 | Task Score (%) | 93 |
| OpenVLM MMT-Bench - Gaze Estimation | 35 | Score (%) | 93 |
| MEGA-Bench Task - MFC Bench Check Face Attribute Edit | 57.1 | Task Score (%) | 92.3 |
| OpenVLM MMT-Bench - Order Hallucination | 55 | Score (%) | 92.2 |
| OpenVLM MMT-Bench - Polygon Localization | 65 | Score (%) | 91.7 |
| OpenVLM MMT-Bench - Sign Language Recognition | 45 | Score (%) | 91.7 |
Interactive version: theaggregate.ai/model?slug=phi-3-5-vision · How It Works · Data refreshed daily, snapshot 2026-11-08.