Loading…
Loading…
Models / Kimi-VL-A3B
1 model (~16.4B) · ↓ 164.9K downloads a month · from $0.46/h on RTX A6000 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/Kimi-VL-A3B-Instruct | vLLM | 33 GB | 40.7 GB | $0.46/h RTX A6000 |
| Kimi VL A3B Thinkingmoonshotai/Kimi-VL-A3B-Thinking | vLLM | 33 GB | 40.7 GB | $0.46/h RTX A6000 |
| Kimi VL A3B Thinking 2506moonshotai/Kimi-VL-A3B-Thinking-2506 | vLLM | 33 GB | 40.7 GB | $0.46/h RTX A6000 |
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").