Local model setup
Preparing Gemma on this device
WAVE uses Gemma locally for sessions, check-ins, reflections, and insights. The first visit downloads and caches the model; after that, the app can reuse it from browser storage.
Ready to startchecking device support
The download has not started yet.
- Model
- onnx-community/gemma-4-E2B-it-ONNX
- Current file
- checking cache
Start when you're ready to use a model-backed page. Chrome or Edge with WebGPU gives the smoothest demo; cached loads should be much faster.