Phi-4
Microsoft's 14B model, strong at maths and reasoning for its size.
Overview
Phi-4 is a 14B model from Microsoft's Phi family, trained largely on synthetic and carefully filtered data. It's strong at maths, reasoning and coding for its size, with a 16K-token context window, and it fits on a single consumer GPU when quantized. It's released under the MIT licence.
| Weights | Open |
|---|---|
| Parameters | 14.7B |
| Context window | 16K tokens |
| Licence | MIT |
| Commercial use | Yes |
LLM hosts that offer Phi-4
All hosts →| LLM host | Hosting plan | Hosting type | Pricing |
|---|---|---|---|
| xCloud | xCloud LocalAI Hosting | GPU cloud | From US$59.99Monthly subscription |
| xCloud | xCloud Ollama Hosting | GPU cloud | From US$59.99Monthly subscription |
News about this model
All news →xCloud joins LLM Hosting Pro with managed Ollama and LocalAI plans
We've added xCloud as a host, with two managed CPU plans for running open-weight models on Ollama or LocalAI. We've also added Microsoft's Phi family to the site, starting with Phi-4.

