# Local. Private. Use Phi-4 in Microsoft Word.

Looking for an alternative to [Microsoft Copilot in Word](https://support.microsoft.com/en-us/microsoft-365-copilot) without recurring inference costs? Consider using local LLMs directly within Microsoft Word. For example, [Mehul’s post](https://medium.com/data-science-in-your-pocket/microsoft-phi-4-the-best-smallest-llm-1cbaa5706e9e) highlights that [Phi-4](https://huggingface.co/microsoft/phi-4) is currently considered the leading choice among small-sized LLMs due to its significant improvements over previous versions. Therefore, it’s worth experimenting with integrating Phi-4 into Microsoft Word to explore its capabilities. You can view a demonstration video showcasing how seamlessly Phi-4 integrates into Microsoft Word using [LM Studio](https://lmstudio.ai/). The most appealing aspect of this method is that it incurs no ongoing inference costs and functions as long as your computer remains operational.

Here’s a quick demo. The demo is powered by [**GPTLocalhost**](https://gptlocalhost.com/), which offers the same core features for individual use. [**LocPilot for Word**](https://locpilot.com/) is the Intranet edition of [**GPTLocalhost**](https://gptlocalhost.com/) designed for enterprise users and team collaboration.

%[https://youtu.be/vL8ND13DNMc] 

For more creative uses of local and private LLMs in Microsoft Word, explore additional demos available on our channel at [**@LocPilot**](https://www.youtube.com/@LocPilot). We hope this demonstration sparks new ideas and encourages you to create further applications that equip your team with cutting-edge on-premise AI capabilities. If you have any particular suggestions or concepts, please feel free to reach out to us at [info@locpilot.com](mailto:info@locpilot.com). Our mission is to make local AI more accessible and affordable for teamwork.