No prompt history yet. Speak into the mic or type instructions to start co-painting!
Choose speech transcription (ASR) and natural neural speech (TTS)
Real-time cloud streaming ASR with sub-second latency, human-grade accuracy, and smart turn detection.
Runs 100% in-browser via WebGPU or WASM SIMD. No API keys needed, zero audio leaves your machine.
Built-in browser speech recognition (Chrome/Safari/Edge). Simple and lightweight.
Ultra-realistic, studio-quality neural voices. Natural inflections and human prosody with built-in acoustic reverb isolation.
Follow these quick steps to enable speech on your device
Mobile browsers block microphone access on insecure local network addresses (http://). Please open via the secure tunnel:
Nice idea, it'll add to the beauty!