Compare LLM Models
Select 2-5 models to compare side by side.
Select models to compare
0 / 5 selected
Top downloads
Select 2-5 models
| Attribute |
qwen
|
|---|---|
| Company | qwen |
| Release Date | 2026-02-25 |
| Parameters (B) | 122 |
| Architecture | Gated DeltaNet + Sparse MoE (hybrid attention) |
| Context Window | 262144 |
| Input Modalities | text, image |
| Open Source | No |
| License | |
| Score | 18 |
| VRAM (GB) | 640 |
| Compute | 8x NVIDIA GPU 80GB (tensor parallel, 262K context) |
| Benchmark: BrowseComp | 63.8 |
| Benchmark: C-Eval | 91.9 |
| Benchmark: CodeForces | 2100 |
| Benchmark: GPQA-Diamond | 86.6 |
| Benchmark: HLE | 25.3 |
| Benchmark: IFEval | 93.4 |
| Benchmark: LiveCodeBench | 78.9 |
| Benchmark: MMLU-Pro | 86.7 |
| Benchmark: MMMU | 83.9 |
| Benchmark: MMMU-Pro | 76.9 |
| Benchmark: SWE-Bench-Verified | 72 |
| Benchmark: Terminal-Bench-2.0 | 49.4 |
| Pricing (Input) | 0.26 |
| Pricing (Output) | 2.08 |