Models / Qwen2
Qwen2 by Qwen
6 models (0.5B to 72B) · ↓ 1.5M downloads a month · from $0.10/h on 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.
Qwen2 0.5B
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|
| OriginalQwen/Qwen2-0.5B | vLLM | 990 MB | 3.4 GB | $0.10/h RTX 3060 |
| InstructQwen/Qwen2-0.5B-Instruct | vLLM | 990 MB | 3.4 GB | $0.10/h RTX 3060 |
Qwen2 1.5B
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|
| OriginalQwen/Qwen2-1.5B-Instruct | vLLM | 3.1 GB | 5.9 GB | $0.10/h RTX 3060 |
| AWQQwen/Qwen2-1.5B-Instruct-AWQ | vLLM | 1.6 GB | 4.3 GB | $0.10/h RTX 3060 |
Qwen2 1.5B
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|
| OriginalQwen/Qwen2-1.5B | vLLM | 3.1 GB | 5.9 GB | $0.10/h RTX 3060 |
Qwen2 7B
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|
| OriginalQwen/Qwen2-7B | vLLM | 15 GB | 19.7 GB | $0.18/h 2x RTX 3060 |
| InstructQwen/Qwen2-7B-Instruct | vLLM | 15 GB | 19.7 GB | $0.18/h 2x RTX 3060 |
Qwen2 57B
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|
| OriginalQwen/Qwen2-57B-A14B | vLLM | 115 GB | 130.3 GB | $2.36/h 4x Nvidia A40 |
| InstructQwen/Qwen2-57B-A14B-Instruct | vLLM | 115 GB | 130.3 GB | $2.36/h 4x Nvidia A40 |
Qwen2 72B
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|
| OriginalQwen/Qwen2-72B | vLLM | 145 GB | 166.8 GB | $1.38/h 8x RTX 3090 |
| InstructQwen/Qwen2-72B-Instruct | vLLM | 145 GB | 166.8 GB | $1.38/h 8x RTX 3090 |
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").