sauravpanda/BrowserAI

Run local LLMs like llama, deepseek-distill, kokoro and more inside your browser

What it solves

BrowserAI provides a way to run production-ready AI models directly in the web browser, eliminating the need for expensive server-side infrastructure and ensuring 100% user privacy by keeping all data processing local.

How it works

It uses a JavaScript SDK that leverages WebGPU acceleration for near-native performance. The library supports multiple inference engines—including MLC, Transformers, Flare (for GGUF models via WASM), and Demucs—allowing developers to load and run a variety of pre-optimized models for text, speech, and audio tasks.

Who it’s for

  • Web developers building AI-powered applications.
  • Companies requiring privacy-conscious AI solutions.
  • AI researchers and hobbyists wanting to experiment without infrastructure overhead.
  • No-code platform builders.

Highlights

  • Local Execution: All processing happens in the browser, enabling offline capability and zero server costs.
  • Multimodal Capabilities: Supports text generation (LLMs), speech-to-text (Whisper), text-to-speech (Kokoro), and audio source separation (Demucs).
  • Developer-Friendly API: Simple SDK for loading models, generating text with JSON schema support, and managing conversations.
  • Broad Model Support: Compatible with Llama, Qwen, Gemma, Phi, and DeepSeek models across different engines.

Related

  • Project
    mlc-ai/web-llm

    A high-performance in-browser LLM inference engine that uses WebGPU to run language models locally without a server.

  • Project
    Ryan-yang125/ChatLLM-Web

    A private, browser-native AI studio and agent workspace that runs LLMs locally on your device using WebGPU, requiring no API keys or backend servers.

  • Project
    addyosmani/chatty

    A private, in-browser AI chat interface that uses WebGPU to run LLMs locally on your hardware for offline use and data privacy.

  • Project
    webmachinelearning/prompt-api

    A proposed web platform API that gives developers a direct, standardized way to access built-in browser or OS language models for local, private, and cost-effective AI integration.

  • Project
    ngxson/wllama

    A WebAssembly binding for llama.cpp that enables running LLMs, multimodal models, and tool calling directly in the web browser using WebGPU and Wasm.