Loading…
Loading…
Models / granite
2 models (3B to 8B) · ↓ 44.3K 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.
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalibm-granite/granite-3b-code-base-2k | vLLM | 7.0 GB | 10.7 GB | $0.10/h RTX 3060 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| granite 8b code instruct 4kibm-granite/granite-8b-code-instruct-4k | vLLM | 16 GB | 20.8 GB | $0.19/h 2x 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").