Loading…
Loading…
Models / CodeLlama
3 models (7B to 13B) · ↓ 13.4K downloads a month · from $0.16/h on RTX 3080 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 |
|---|---|---|---|---|
| Originalmeta-llama/CodeLlama-7b-Python-hfLicence to accept | vLLM | 13 GB | ~18.4 GB | $0.16/h RTX 3080 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalmeta-llama/CodeLlama-7b-hfLicence to accept | vLLM | 13 GB | ~18.4 GB | $0.16/h RTX 3080 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalmeta-llama/CodeLlama-13b-Instruct-hfLicence to accept | vLLM | 26 GB | ~33.5 GB | $0.35/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").