Loading…
Loading…
Models / Qwen2.5-Coder
1 model (14B) · ↓ 2.1M downloads a month · from $0.14/h on RTX A4000 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 |
|---|---|---|---|---|
| InstructQwen/Qwen2.5-Coder-14B-Instruct | vLLM | 30 GB | 36.8 GB | $0.45/h Nvidia A40 |
| Instruct AWQQwen/Qwen2.5-Coder-14B-Instruct-AWQ | vLLM | 10.0 GB | 15.1 GB | $0.14/h RTX A4000 |
| Instruct GPTQ Int4Qwen/Qwen2.5-Coder-14B-Instruct-GPTQ-Int4 | vLLM | 10.0 GB | 15.1 GB | $0.14/h RTX A4000 |
| Instruct bnb 4bit (unsloth)unsloth/Qwen2.5-Coder-14B-Instruct-bnb-4bit | vLLM | 9.9 GB | 15 GB | $0.14/h RTX A4000 |
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").