Gemma Lab

Translate text with Gemma 4 running right here in your browser. The Lab's WASM engine runs kapi translate --provider gemma, and the model itself executes via transformers.js on WebGPU — the same ONNX model the native kapi-llm plugin runs in-process. It is a free, private alternative to the paid cloud providers: the model is a one-time multi-GB download (then cached by your browser), and nothing leaves the page. A WebGPU-capable browser is required.