Eleven Labs Podcast

Quick Usage (Already Configured)

1) List available voices

bash .opencode/skills/eleven-labs-podcast/scripts/list-voices.sh

2) Generate podcast audio from a text file

python3 .opencode/skills/eleven-labs-podcast/scripts/create-podcast.py "notes/episode-01.txt" --mode auto --output "notes/episode-01.mp3"

3) Force dialogue mode (HOST/GUEST)

python3 .opencode/skills/eleven-labs-podcast/scripts/create-podcast.py "notes/dialogue.txt" --mode dialogue

Input Format

• auto mode uses dialogue when it sees HOST: or GUEST: lines; otherwise it uses narration (tts).
• tts mode narrates the full file with one voice.
• dialogue mode treats each non-empty line as one turn.
• In dialogue mode, unprefixed lines alternate between host and guest voices.

What This Skill Loads

• Credentials are loaded from .opencode/skills/eleven-labs-podcast/.env.
• Required: ELEVEN_LABS_API.
• Optional defaults come from .env.example.
• API calls used by this skill:
  • GET /v1/voices
  • POST /v1/text-to-speech/{voice_id}
  • POST /v1/text-to-dialogue

Common Gotchas

• If you get a 401 or 403, verify the API key and key scope.
• mp3_44100_192 requires a higher ElevenLabs tier; default is mp3_44100_128.
• For dialogue mode, split very long lines into shorter turns for best quality.

First-Time Setup (If Not Configured)

1. Copy .opencode/skills/eleven-labs-podcast/.env.example to .opencode/skills/eleven-labs-podcast/.env.
2. Set ELEVEN_LABS_API.
3. Optionally set your preferred voice IDs and model IDs.