Mantis-8B-clip-llama3: benchmark results
Provider: Salesforce. Released 2024-05-29. Access: Open.
Aggregate Rating 1301 ± 35, rank #8733 of 10916 rated models, from 439 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMT-Bench - Spot The Diff | 100 | Score (%) | 99.5 |
| OpenVLM MMBench V1.1 EN - Image Topic | 98.9 | Accuracy (%) | 95.2 |
| OpenVLM MMT-Bench - Transparent Object Detection | 80 | Score (%) | 92.7 |
| OpenVLM MMT-Bench - Artwork Emotion Recognition | 60 | Score (%) | 91.7 |
| OpenVLM MMBench V1.1 CN - Image Topic | 98.9 | Accuracy (%) | 90.3 |
| OpenVLM MMT-Bench - Micro Expression Recognition | 40 | Score (%) | 89.8 |
| OpenVLM A-Bench - Contradiction Overcome | 87.5 | Accuracy (%) | 89 |
| OpenVLM MMT-Bench - MIA | 70 | Score (%) | 88.6 |
| OpenVLM MMT-Bench - Temporal Sequence Understanding | 50 | Score (%) | 88.1 |
| OpenVLM MMT-Bench - Gesture Recognition | 70 | Score (%) | 86.9 |
| OpenVLM Q-Bench - Yes-or-No (Others) | 77.7 | Accuracy (%) | 85.4 |
| OpenVLM MMT-Bench - Animated Character Recognition | 60 | Score (%) | 85 |
Interactive version: theaggregate.ai/model?slug=mantis-8b-clip-llama3 · How It Works · Data refreshed daily, snapshot 2026-11-08.