Multi-voice dialogue demo: Five vocal personalities for the same script

Use the dialogue array of ElevenLabs text-to-dialogue to construct a two-person short play; by assigning a different voice_id to each line, the same scene is split into dialogue layers with different vocal personalities such as steady, bright, and deep. Set stability to a balanced value to preserve intonation variation, and leave language_code empty so the model detects it automatically. This kind of structure is suitable for developers who, before integrating conversational speech, want to quickly verify whether multi-speaker voice pairings, line pacing, and emotional expression fit narration, podcast, or voice agent scenarios in their product. The work is genuinely generated by the author's pipeline, and the voice field uses the preset voice ids listed in the schema.