Orpheus v1
Orpheus v1A fast, expressive text-to-speech model with six distinct voices, low-latency audio, and bracketed controls for tone, style, and delivery.
Model details
Orpheus v1 is Canopy Labs’ newest expressive text-to-speech model (successor to Orpheus TTS) for generating fast, high-quality audio with low-latency inference . It offers six professionally trained voices and supports bracketed directions such as cheerful, whisper, or dramatic to control tone and delivery. Without directions, it produces a more natural, conversational cadence.
The model is suited to AI assistants, customer support, game characters, narration, podcasts, audiobooks, and other expressive content.
Input/Output
Canopy Labs recommends maximum 200 characters per request (see Orpheus v1 Prompting Guide). Orpheus v1 generates up to 30 seconds of audio.
Supported Languages
English, French, German, Hindi, Italian, Japanese, Korean, Mandarin, Portuguese, Russian, Spanish
