ON-DEVICE AI · NO ACCOUNT NEEDED
See it. Say it.
Learn it.
Turn everyday objects into words.
Choose a language. Tap an object. Hear it spoken.
A new word,
right in front of you.
Choose an object
Tap an objectThe first scan downloads a model. No camera permission is needed for the sample.
Detection runs in your browser. Images are not uploaded. Sample credit
Choose an object
—
Your next word starts with a photo.
Tap a detection box or an object name.
Voice & detection settings
How it works & limitations
Try the sample photo or open your camera, then tap a detection box or an object name. Change the word language and press the speaker button to hear it. The sample uses the real detector, not preset boxes.
A selected word stays visible when its object leaves the frame. Tracking may confuse objects of the same type. This detects object categories; it does not translate text or identify specific products.
Pronunciation uses your browser's speech voices. Availability and sound vary by device. Some voices use an online speech service. Object words are provided by a curated vocabulary, not generated translations.