Microsoft · phi3
Can I run Phi-3 Mini?
Microsoft's Phi-3 Mini — compact 3.8B model with strong reasoning. Supports up to 128K context.
3.8B parameters128K contexttextMIT1 builds
Phi-3 Mini system requirements
Memory needed at 4K context, counting model weights, KV cache, compute buffers and runtime overhead. Download sizes come from the official Ollama registry.
| Build | Quantization | Download | Memory needed |
|---|---|---|---|
| phi3:mini | Q4_K_M | 2.3 GB | ~4.6 GB |
How to run Phi-3 Mini locally
- 1.Install Ollama for Windows, macOS or Linux.
- 2.Run this in a terminal:
ollama run phi3:mini - 3.The weights download on first run, then an interactive prompt opens.
Frequently asked questions about Phi-3 Mini
- How much VRAM do I need to run Phi-3 Mini?
- At 4K context, the smallest published build of Phi-3 Mini (Q4_K_M) needs roughly 4.6 GB of GPU memory once weights, KV cache, compute buffers and runtime overhead are counted. Bigger quantizations and longer context windows need more. It can also run partly or entirely on the CPU using system RAM, more slowly.
- Can I run Phi-3 Mini without a dedicated GPU?
- Yes, but slowly. Without a GPU the model runs on the CPU using system RAM, which usually means a few tokens per second rather than dozens. You would need at least 4.6 GB of free RAM for the smallest build.
- How big is the Phi-3 Mini download?
- The smallest published build is 2.3 GB. Leave some extra free disk space beyond the download itself.
- How do I run Phi-3 Mini locally?
- Install Ollama, then run "ollama run phi3:mini" in a terminal. The weights download on first use and an interactive session opens.
Will Phi-3 Mini run on your PC?
Scan your hardware once and get a verdict for this model and every build — free, no account.
Model data from the official Ollama registry · verified 13/09/2026