Orpheus TTS Podcast Generator

Orpheus TTS Podcast Generator uses Canopy AI's Orpheus TTS and Gemini to create high-quality, multi-speaker podcasts.
Generate dynamic, engaging, and natural-sounding conversations in multiple languages.

Features

• Multi-Speaker Support: Generate conversations between multiple characters with distinct voices.
• Gemini Integration: Utilizes Gemini to create podcast scripts and conversational content.
• Emotive Speech: Orpheus TTS supports emotive tags for realistic and engaging delivery.
• High-Quality Audio: Produces clear and natural-sounding audio at 24kHz.
• Multi-Language Support: English, German, French, Spanish, Italian, Korean, Hindi, Chinese.

Model Selection: Finetuned vs Pretrained

You can choose between finetuned and pretrained Orpheus TTS models for each language.
Finetuned models are recommended for best quality and are the default. You can switch to pretrained models using the --model-type argument or the PODCAST_MODEL_TYPE environment variable.

Live Example Output

How It Works

• Setup: Install dependencies from requirements.txt.
• Configure: Set your desired voices and Gemini prompts in the config or via CLI.
• Generate: Run the script to create your podcast audio and download the result.