SenseNova: benchmark results
Provider: SenseTime. Released 2024-04-24. Access: API.
Aggregate Rating 1594 ± 26, rank #1080 of 10916 rated models, from 230 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMMU - Public Health | 96.7 | Accuracy (%) | 99.3 |
| OpenVLM MMBench V1.1 CN - Physical Relation | 76 | Accuracy (%) | 98.9 |
| OpenVLM MMStar - Science and Technology | 70 | Accuracy (%) | 98.9 |
| OpenVLM MMBench V1.1 EN - Physical Relation | 76 | Accuracy (%) | 98.6 |
| OpenVLM MMMU - Psychology | 83.3 | Accuracy (%) | 98.6 |
| OpenVLM MathVista - FQA | 80.7 | Accuracy (%) | 98.3 |
| OpenVLM MathVista - MWP | 89.8 | Accuracy (%) | 98 |
| OpenVLM MathVista - STA | 88.4 | Accuracy (%) | 98 |
| OpenVLM MMStar - Math | 85.2 | Accuracy (%) | 97.9 |
| OpenVLM MMBench V1.1 EN - Object Localization | 91.4 | Accuracy (%) | 97.7 |
| OpenVLM MMStar - Logical Reasoning | 79.6 | Accuracy (%) | 97.3 |
| OpenVLM MM-Vet - Math | 91.5 | Score | 97 |
Interactive version: theaggregate.ai/model?slug=sensenova · How It Works · Data refreshed daily, snapshot 2026-11-08.