Qwen 4 Max (Leaked Staging) vs Visual Jev (Qwen3-VL 8B Backbone)
A direct, empirical evaluation of Qwen 4 Max (Leaked Staging) (Alibaba Cloud) against Visual Jev (Qwen3-VL 8B Backbone) (Independent AI Systems Lab (Guanxu Yu & Yuhang Yao)).
Qwen 4 Max (Leaked Staging)
Developed by Alibaba Cloud · 1.2 Trillion (Sparse MoE) parameters
Repository-scale software engineering outscoring GPT-6 Sol, multilingual enterprise agents across 120 languages, and low-cost 1M-token processing.
Visual Jev (Qwen3-VL 8B Backbone)
Developed by Independent AI Systems Lab (Guanxu Yu & Yuhang Yao) · 8.2 Billion Vision-Language Parameters parameters
Sub-20ms multi-query image verification, automated rejection sampling in generative diffusion pipelines (FLUX / SD3), robotics perception routing, and high-frequency visual software inspection.
| Metric | Qwen 4 Max (Leaked Staging) | Visual Jev (Qwen3-VL 8B Backbone) |
|---|---|---|
| SWE-bench Verified | 78.6% | 62.1% |
| MMLU-Pro | 91.4% | 86.4% |
| MATH-500 | 95.6% | 89.2% |
| Context Window | 1,000,000 tokens | 32,768 tokens (Visual prefix + batched question suffixes) |
| Input Pricing (per 1M) | $1.20 | $0.05 |
| Output Pricing (per 1M) | $4.80 | $0.10 |