abab7-preview: benchmark results
Provider: MiniMax. Released 2024-11-06. Access: API.
Aggregate Rating 1575 ± 33, rank #1329 of 10916 rated models, from 433 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMT-Bench - Instance Captioning | 100 | Score (%) | 99.8 |
| OpenVLM MMMU - Music | 50 | Accuracy (%) | 99.5 |
| OpenVLM MMT-Bench - Health and Medicine | 92.9 | Score (%) | 99 |
| OpenVLM MMT-Bench - Counting By Reasoning | 95 | Score (%) | 98.8 |
| OpenVLM MME - Color | 195 | Score | 98.7 |
| OpenVLM MMT-Bench - Image Quality Assessment | 75 | Score (%) | 98.5 |
| OpenVLM LLaVA-Bench - Conv | 101.4 | Score | 98.3 |
| OpenVLM MMT-Bench - Religious Recognition | 85 | Score (%) | 98.3 |
| OpenVLM MMT-Bench - Tech Engineering | 61.1 | Score (%) | 98.3 |
| OpenVLM MMMU - Diagnostics and Laboratory Medicine | 56.7 | Accuracy (%) | 98.2 |
| OpenVLM MMT-Bench - Weapon Recognition | 100 | Score (%) | 98.1 |
| OpenVLM MMMU - Management | 76.7 | Accuracy (%) | 97 |
Interactive version: theaggregate.ai/model?slug=abab7-preview · How It Works · Data refreshed daily, snapshot 2026-11-08.