Mammoth-VL: benchmark results
Provider: CMU. Access: Open.
Aggregate Rating 1461 ± 27, rank #3279 of 10916 rated models, from 545 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MEGA-Bench Task - Red Teaming Captcha | 16.3 | Task Score (%) | 100 |
| MEGA-Bench Task - Red Teaming Visual Order B | 91.6 | Task Score (%) | 100 |
| MEGA-Bench Task - Rocks Samples Compare | 78.6 | Task Score (%) | 100 |
| MEGA-Bench Task - Red Teaming Visual Misleading | 91.1 | Task Score (%) | 98.8 |
| MEGA-Bench Task - Pictionary Cartoon Drawing Guess | 85.7 | Task Score (%) | 96.5 |
| MEGA-Bench Task - ActivityNetQA | 54.2 | Task Score (%) | 95.3 |
| MEGA-Bench Task - Bridge Strategies Expert | 45 | Task Score (%) | 95.3 |
| MEGA-Bench Task - CLEVRER Object Existence Video | 87.5 | Task Score (%) | 95.3 |
| MEGA-Bench Task - Egocentric Spatial Reasoning | 66.7 | Task Score (%) | 91.9 |
| MEGA-Bench Task - KVQA Knowledge Aware QA | 57.9 | Task Score (%) | 91.9 |
| MEGA-Bench Task - Research Website Parsing Blog Post | 21.4 | Task Score (%) | 91.9 |
| MEGA-Bench Task - Trance Physics Reasoning Basic | 88.2 | Task Score (%) | 90.7 |
Interactive version: theaggregate.ai/model?slug=mammoth-vl · How It Works · Data refreshed daily, snapshot 2026-11-08.