AI · Web app

TranslateGemma 4B WebGPU

Private, in-browser machine translation running Google's TranslateGemma model via WebGPU

Visit huggingface.co ↗

TranslateGemma 4B WebGPU is a browser-based translation demo powered by Google's TranslateGemma model and Transformers.js. All processing runs client-side using WebGPU and ONNX Runtime Web, meaning translations are completely private and work offline once the model finishes downloading. It supports translation across 56 languages without sending text to an external server.

What it's for

  • Translate confidential or sensitive text without sending data to third-party servers
  • Perform offline machine translation in the browser without an active internet connection
  • Test on-device WebGPU inference capabilities with large language models in browser environments

Features

  • Client-side inference using Transformers.js and ONNX Runtime Web
  • Hardware acceleration via WebGPU
  • Privacy-focused architecture with zero server-side data transfer
  • Offline functionality once initial model assets are cached
  • Support for translation across 56 languages

Pricing

Free to use directly in the browser

Details

Type
Web app
Category
AI
Topic
Machine Translation
Open source
Yes
Platforms
Web
For
Developers, researchers, and privacy-conscious users needing local text translation
Alternative to
Google Translate, DeepL
Tags
translation, webgpu, transformers js, onnx, local ai, nlp

Why it was saved

  • @nicodotdev, 24 February 2026: Free, private, offline-capable translation across ~55 languages with no server and no data sent anywhere; model is a 2.9GB download and requires a WebGPU-capable browser with sufficient VRAM.

Related

  • TranslateGemma (google/translategemma-4b-it)

    Open translation model by Google DeepMind lightweight enough for on-device use.

  • HeyGen

    AI avatar video generation and translation platform driven by text prompts

  • Pinokio

    Browser that installs and runs open-source AI apps with one click

  • Local Deep Researcher

    Fully local web research and report writing assistant using Ollama or LMStudio

  • AgenticSeek

    Open-source, fully local autonomous AI agent that browses and writes code

  • AnythingLLM

    Private, on-device AI assistant to chat with documents and local models.

Updated