# Run local AI on Windows

Use Ollama or LM Studio for a local chat on Windows; the Hugging Bay Windows app and signed native hbay installer are not yet published.

Sources checked: 2026-10-17

## 1. Check your PC

Open Settings → System → About. Note Installed RAM, System type and Windows version. Use [the memory guide](/learn/choose-a-model) to choose a small first model. A dedicated GPU can help, but VRAM is separate from system RAM.

## 2. Choose LM Studio for a graphical app

Open [LM Studio downloads](https://lmstudio.ai/download) and choose the Windows build for your system. The official requirements list x64 with AVX2 or supported ARM systems and recommend at least 16 GB RAM. Run the installer, then open LM Studio.

Follow [the LM Studio guide](/learn/lm-studio): Discover, download a small compatible model, open Chat, load it and send a message.

## 3. Or choose Ollama

The official Windows instructions require Windows 10 22H2 or later. Download the Windows installer from [Ollama](https://ollama.com/download), run it, and open Ollama. Open a new PowerShell window and check `ollama --version`.

Follow [the Ollama guide](/learn/ollama) for the first model. You do not need WSL for the native Windows app. If you want GPU acceleration, check the current driver requirements linked below.

## Using Hugging Bay on Windows

Use [the website](/finder) to find model files and read their licenses. A compatible downloaded GGUF can be imported into LM Studio or Ollama using their guides.

On 17 October 2026, [our desktop page](/desktop) listed Windows as coming soon and [our install page](/install) did not publish a signed native Windows installer. Use one of the supported runners above.

## Sources

- [Hugging Bay desktop downloads](https://huggingbay.xyz/desktop)
- [Hugging Bay install status](https://huggingbay.xyz/install)
- [Ollama on Windows](https://docs.ollama.com/windows)
- [LM Studio system requirements](https://lmstudio.ai/docs/app/system-requirements)

[Next: chat with LM Studio](/learn/lm-studio)