Gemini 4 Pro (Leaked Benchmark) vs Visual Jev (Qwen3-VL 8B Backbone)
A direct, empirical evaluation of Gemini 4 Pro (Leaked Benchmark) (Google DeepMind) against Visual Jev (Qwen3-VL 8B Backbone) (Independent AI Systems Lab (Guanxu Yu & Yuhang Yao)).
Gemini 4 Pro (Leaked Benchmark)
Developed by Google DeepMind · Proprietary Omni-Modal Unified Diffusion Transformer parameters
Repository-scale 2M-token refactoring (88.7% DeepSWE v1.1), headless shell execution (95.3% Terminal-Bench 2.1), and multimodal desktop agent autonomy (86.8% OSWorld 2.0).
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 | Gemini 4 Pro (Leaked Benchmark) | Visual Jev (Qwen3-VL 8B Backbone) |
|---|---|---|
| SWE-bench Verified | 88.7% | 62.1% |
| MMLU-Pro | 93.8% | 86.4% |
| MATH-500 | 97.8% | 89.2% |
| Context Window | 2,000,000 tokens | 32,768 tokens (Visual prefix + batched question suffixes) |
| Input Pricing (per 1M) | $2.25 | $0.05 |
| Output Pricing (per 1M) | $11.25 | $0.10 |