audiospeech
Orpheus TTS
Llama-based Orpheus TTS with 8 empathetic voices for agents and audiobooks.
Studio is in early access; availability resolves per project after sign-in.
- Endpoints
- 1
- Eligible
- 1
- Input schemas imported
- 1 / 1
- Delivery
- fal.ai
- Weights
- Not stated
- Developer
- Not stated in source
Overview
Orpheus TTS delivers human-level speech synthesis on a finetuned Llama-based Speech-LLM, prioritizing prosody and clarity. Eight voices (Tara, Leah, Jess, Leo, Dan, Mia, Zac, Zoe) cover conversational range. Voice agents, audiobooks, and accessibility tools lead uses.
Capabilities
- Llama-based Speech-LLM
- 8 distinct voices
- Empathetic prosody
- Human-level clarity
Best for
- Voice agents
- Audiobook narration
- Accessibility tools
Use cases
- Conversational AI
- Screen readers
- Expressive narration
Endpoints
1 eligible. Cataloged is not the same as executable: Studio resolves which endpoints a project can run.
| Endpoint | Task | Catalog status | Inputs |
|---|---|---|---|
fal-ai/orpheus-tts | text to audio | Eligible | 4 fields · requires text |
Questions
Architecture?
Llama-based Speech-LLM.
Voices?
8 named voices.