SenseNova-V6-5-Pro: benchmark results
Provider: SenseTime. Released 2025-07-27. Access: API.
Aggregate Rating 1691 ± 22, rank #389 of 10916 rated models, from 161 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MM-Vet | 89.4 | Score | 100 |
| OpenVLM HallusionBench - qAcc | 60.7 | Accuracy (%) | 100 |
| OpenVLM Leaderboard | 82.2 | Avg Score | 100 |
| OpenVLM MM-Vet - Knowledge | 85.7 | Score | 100 |
| OpenVLM MM-Vet - Language Generation | 86.4 | Score | 100 |
| OpenVLM MM-Vet - Recognition | 89.3 | Score | 100 |
| OpenVLM MMBench V1.1 CN - Nature Relation | 97.8 | Accuracy (%) | 100 |
| OpenVLM MMBench V1.1 EN - Nature Relation | 97.8 | Accuracy (%) | 100 |
| OpenVLM MMMU - Business | 90.7 | Accuracy (%) | 100 |
| OpenVLM MMMU - Energy and Power | 83.3 | Accuracy (%) | 100 |
| OpenVLM MMStar - Science and Technology | 76.8 | Accuracy (%) | 100 |
| OpenVLM MathVista - FQA | 82.5 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=sensenova-v6-5-pro · How It Works · Data refreshed daily, snapshot 2026-11-08.