Route AI model requests through a unified proxy endpoint
Chat with an AI assistant using adjustable settings
Chat with Gemma 4 E2B AI model in your browser
Run LFM2.5 230M with custom WebGPU kernels