A browser-based chat application with AI models that run entirely on your device without sending data to external servers. This project uses Svelte 5, SvelteKit, WebAssembly, and the Wllama library.
You can try the application at: https://svelte-local-ai.khromov.se/
Clone the repository
Install dependencies
nvm use
npm install
Start the development server
npm run dev
Open your browser and navigate to http://localhost:5173
npm run build
# Build the Docker image
docker build -t sveltekit-local-ai .
# Run the container
docker run -p 3000:3000 sveltekit-local-ai
The application downloads compact language models directly to your browser and runs inference using WebAssembly. This approach ensures your conversations stay private and can work offline after the initial model download.