Loading…
Loading…
Models / Moonlight
2 models (16B to 16B) · ↓ 53.4K downloads a month · from $0.42/h on 4x RTX 3060 right now. Each model's versions do the same job; quantized ones (AWQ, FP8, 4-bit) need less GPU memory and run on cheaper machines.
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalmoonshotai/Moonlight-16B-A3B-Instruct | vLLM | 32 GB | 39.7 GB | $0.42/h 4x RTX 3060 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalmoonshotai/Moonlight-16B-A3B | vLLM | 32 GB | 39.7 GB | $0.42/h 4x RTX 3060 |
Model details come from Hugging Face and are refreshed daily; prices follow the live offers. GPU memory is GridPup's estimate: the weights, room for an 8,192-token conversation (chat models) and the runtime's own memory. Large models are split across the GPUs of one machine ("4x RTX 4090").