Generate Podcast Episodes with Multiple AI Hosts

Give the agent a topic and a few bullet points. It drafts a dialogue script, renders each host in a distinct ElevenLabs voice using dialogue mode, stitches the audio, and delivers a finished episode.

WhatsApp DMpodcastdialoguecontent

Conversation example

Generate a 6-minute podcast episode: two hosts debating whether AI agents will replace SaaS dashboards. Host A is skeptical, Host B is bullish. Casual tone.

task_plan

Drafting the dialogue script — 14 exchanges, roughly 6:00 total...

audio_generate

Script ready. Rendering with ElevenLabs dialogue mode: - Host A → voice_id=rachel_analyst (skeptical) - Host B → voice_id=adam_builder (bullish) Using Multilingual v2 with natural turn-taking and breath markers.

publish_file

Episode ready: File: agents-vs-dashboards-ep01.mp3 (5.8 MB) Duration: 6:12 Speakers: 2 (Rachel, Adam) Characters used: 8,420 Model: Multilingual v2 + dialogue stitching Highlights the agent caught: - 1:12 — good back-and-forth on context windows - 3:45 — Adam's 'dashboards are WhatsApp with extra steps' line is quotable Published to your WhatsApp DM. Want me to also generate a 30-second social cut?

More ElevenLabs use cases

Deploy this in minutes

Create a LunoGen agent, connect ElevenLabs, and start running this workflow from WhatsApp today.