2 voice dialoge | ElevenLabs | Page 1

AI-generated Answer for roykim1212

Here’s how to create a 2‑voice dialogue with ElevenLabs:

Pick two voices and get their voice_id values. You can copy the ID from My Voices in the dashboard, or list voices via GET /v2/voices. (help.elevenlabs.io)
Use the Text to Dialogue API (model_id must be eleven_v3). Example request body:
{
"model_id": "eleven_v3",
"inputs": [
{"text": "Knock knock.", "voice_id": "VOICE_ID_A"},
{"text": "Who’s there?", "voice_id": "VOICE_ID_B"}
]
}
Send to POST /v1/text-to-dialogue. Optional params include output_format, seed, and pronunciation_dictionary_locators. (elevenlabs.io)
For progressive playback, use POST /v1/text-to-dialogue/stream (up to 10 unique voice_ids per request). If you also need timestamps/alignment, use /v1/text-to-dialogue/stream-with-timestamps. (elevenlabs.io)
If you prefer the website (no code), enable Dialogue mode by adding multiple speakers in the Text to Speech UI. (help.elevenlabs.io)

References:

#2 voice dialoge