System Requirements
Check whether your Mac or Windows PC is ready before installing.
Follow the bold button names and numbered steps. Technical names are included only when you need to recognize a model or file inside the app.
macOS
- Apple Silicon Mac: M1 or newer.
- macOS 13 Ventura or newer.
- 8 GB unified memory for the lightest local models.
- 16 GB or more recommended.
- Internet for installation, model downloads, updates, Telegram, or cloud providers.
- Several gigabytes of free disk space for local models.
Local AI uses Apple Metal acceleration. There is no Intel Mac build of Local Waifu 1.7.3.
Windows
- Windows 10 22H2 or Windows 11.
- 64-bit x64 computer.
- 8 GB RAM minimum; 16 GB or more recommended.
- A GPU is optional. Local Waifu can run on CPU, but it will be slower.
- Current GPU drivers are recommended for NVIDIA, AMD, or Intel acceleration.
- Microsoft Edge WebView2 Runtime; the installer can fetch it if required.
Windows may configure an appropriate GPU runtime after first launch. The app can continue on CPU while setup completes; restart when the app reports that acceleration is ready.
Plan disk space before you install models
The application is much smaller than the AI models it can use. Keep at least 8 to 10 GB free for a comfortable lightweight local setup. Larger chat models can require roughly 17 to 24 GB each, and local image generation adds several more gigabytes.
Typical additional downloads include:
- a chat model: approximately 3 to 24 GB depending on size and quantization;
- EmbeddingGemma for semantic memory: approximately 0.6 GB;
- local image weights: approximately 4 to 5 GB or more;
- speech recognition and voice models: hundreds of megabytes to several gigabytes.
These figures are estimates. The model screen shows the current download size and RAM guidance.