Step 1
Sign in front of the camera
Sora tracks body and hand landmarks locally and proposes isolated sign words. No video is sent to the server.
Two people. One clearer conversation.
Sign, review, speak — or listen and read. Every step is shown on screen.
Step 1
Sora tracks body and hand landmarks locally and proposes isolated sign words. No video is sent to the server.
Step 2
Recognized signs become a sentence. Gemini turns those sign words into a short English interpretation for the signer to check and correct.
Step 3
After confirmation, ElevenLabs turns the approved sentence into speech. You stay in control of what is said aloud.
Step 4
When listening is on, ElevenLabs Scribe turns the hearing person’s speech into readable live captions.
Sora’s current recognizer works with supported isolated signs and phrases. It is not unrestricted continuous ASL translation and should not replace a qualified interpreter in medical, legal, emergency, or other high-stakes situations.