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 |
moonshotai
|
|---|---|
| Company | moonshotai |
| Release Date | 2025-04 |
| Parameters (B) | 16 |
| Architecture | MoE Transformer |
| Context Window | 128000 |
| Input Modalities | text, image |
| Open Source | Yes |
| License | mit |
| Score | 0 |
| VRAM (GB) | 32 |
| Compute | 16B MoE VLM in BF16 needs ~32GB VRAM on a single GPU; Q4 quantization runs on ~11–16GB. Supports HuggingFace Transformers and vLLM inference. |
| Benchmark: MMMU | 61.7 |
| Pricing (Input) | — |
| Pricing (Output) | — |