Loading…
Loading…
Models / Falcon3
4 models (1B to 10B) · ↓ 41.8K 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 |
|---|---|---|---|---|
| Originaltiiuae/Falcon3-3B-Instruct | vLLM | 6.5 GB | 10.2 GB | $0.10/h RTX 3060 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Instructtiiuae/Falcon3-7B-Instruct | vLLM | 15 GB | 19.8 GB | $0.19/h 2x RTX 3060 |
| Base (Prince-1)Prince-1/Falcon3-7B-Base | ONNX Runtime GenAI | 15 GB | 19.6 GB | $0.13/h Nvidia P40 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Instructtiiuae/Falcon3-10B-Instruct | vLLM | 21 GB | 26.6 GB | $0.31/h 2x 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").