Phi 3 Models — Hardware Requirements
6 Phi 3 models from Microsoft and the community, from the smallest that runs in 2.3 GB of VRAM up to 41.9B parameters. Every row links to full quantization tables and GPU compatibility.
All Phi 3 Models by Size
| Model | Params | Runs from | Context | Publisher | Quant downloads |
|---|---|---|---|---|---|
| Phi 3.5 Mini Instruct | 3.8B | 2.3 GB | 131K | ||
| Phi 3 Mini 4k Instruct | 3.8B | 2.7 GB | 4K | ||
| Phi 3 Mini 128k Instruct | 3.8B | 2.7 GB | 131K | ||
| Phi 3 Small 8k Instruct | 7.4B | 15.3 GB | 8K | ||
| Phi 3 Medium 4k Instruct | 14.0B | 6.7 GB | 4K | ||
| Phi 3.5 MoE Instruct | 41.9B | 12.1 GB | 131K |
Frequently Asked Questions
- How much VRAM do I need to run a Phi 3 model?
- The smallest Phi 3 model, Phi 3.5 Mini Instruct, runs from 2.3 GB of VRAM at an aggressive quantization. Larger family members need proportionally more — see the table above for every model.
- Which Phi 3 models can I run on a 16 GB GPU?
- 6 of 6 Phi 3 models fit in 16 GB of VRAM at some quantization, including Phi 3.5 Mini Instruct, Phi 3 Mini 4k Instruct, Phi 3.5 MoE Instruct.
- What is the most popular Phi 3 model to run locally?
- Phi 3.5 Mini Instruct is the most downloaded Phi 3 model in local-friendly quantized formats. It runs from 2.3 GB of VRAM.