Custom wake words and fixed voice commands — in a single on-device model. Train "hey kitchen" to activate, "next step" to navigate, "stop timer" to control. No speech-to-text needed. First model free.
You want to add voice activation to your project — a smart home setup, a mobile app, or a Raspberry Pi. But the only options are enterprise SDKs with per-user pricing that kills your margins — or research projects you have to wrangle yourself.
Outspoken fills the gap. Train a custom wake word through our web UI, then download a standard ONNX model for local testing and prototypes. Credit-based training. Commercial deployment licensing available. First Balanced model free.
Your trained ONNX and TFLite models work everywhere ONNX Runtime or TensorFlow Lite runs. No vendor lock-in. No proprietary formats.
openWakeWord models work natively in Home Assistant's voice pipeline. Train your wake word, drop it in, done.
Compatible with HA 2024.x+
View HA docs →Setup guide →Load your .onnx model with the openwakeword library. Raspberry Pi, edge servers, custom hardware.
pip install openwakeword
ONNX Runtime for Python →Python guide →Use onnxruntime-react-native for on-device wake word detection on iOS and Android. No cloud calls.
npm install onnxruntime-react-native
onnxruntime-react-native docs →React Native guide →Run models with onnxruntime-web. Test in our dashboard or integrate into your own web app.
Already working in the Outspoken dashboard
onnxruntime-web docs →ONNX Runtime has official bindings for native apps, game engines, and embedded systems.
Standard ONNX format, no proprietary dependencies
ONNX Runtime bindings →Dedicated Expo module with useWakeWord hook, config plugin, and prebuild support.
expo-outspoken
All models use the standard ONNX format. ~700KB–2MB per model. No proprietary runtime. No vendor lock-in.
Not just "Hey Siri" clones. Outspoken gives you activation words and intent detection in a single model — on-device, private, no speech recognition needed.
Custom activation phrases like "Hey Chef" or "OK Buddy". Always listening, near-zero false positives. Train your own in minutes.
Fixed command detection — "next step", "go back", "start timer". Each intent adds less than 1ms of processing. No speech-to-text needed.
Models run locally via ONNX Runtime. No audio leaves the device. No server calls. No latency. Works offline.
Train in English, Dutch, German, and French — with more coming. No per-language fees. Same credit cost for every language.
Standard .onnx and .tflite models (700KB–2MB). ONNX for desktop, server, and web; TFLite for Android, microcontrollers, and Coral TPU. No vendor lock-in.
Train custom models through our web UI. Upload nothing — we synthesize training data. Your model is ready in ~45 minutes.
Think of it like a button, but activated by voice. Triggers must be predefined — this is not transcription or conversational AI.
No ML expertise needed. No audio samples to collect. Go from idea to working voice control in under an hour.
Choose a wake word and any intents. Type them in — any language, any phrase.
Synthetic voice data, noise augmentation, negative mining. Done in ~45 minutes.
Lightweight models (700KB–2MB) in both formats. Bundle in your app or load dynamically.
Use locally for personal, hobby, and prototypes. Commercial venue, fleet, or customer deployment requires a license.
Choose a wake word and any intents. Type them in — any language, any phrase.
Synthetic voice data, noise augmentation, negative mining. Done in ~45 minutes.
Lightweight models (700KB–2MB) in both formats. Bundle in your app or load dynamically.
Use locally for personal, hobby, and prototypes. Commercial venue, fleet, or customer deployment requires a license.
Want to understand the training parameters? Read the training guide →
Standard credits cover personal, hobby, prototype, and internal evaluation use. Shipping to customers, venues, or fleets requires a separate commercial deployment license.
No credit card needed
Personal, hobby, and prototype use only. Not licensed for commercial deployment.
Get startedFirst Balanced model free · credits never expire
Personal, hobby, prototype, and internal evaluation use. Customer-facing or fleet deployments need a commercial license (below).
Get started for free✓Your first model is on us — no credit card required
Required for shipping wake words to customers, venues, fleets, or resale. Priced per volume — credits and licensing are separate.
All models run on-device. No subscriptions for training credits.
Train a custom wake word with standard ONNX output. Run locally for personal, hobby, and prototype use. Commercial deployment licensing available. First Balanced model free.