Automatic Content Recognition
Dedicated Account Manager
Early Access to New Features
Emotion Tags
normalslowcryingsleepysighchuckle
Enterprise Language Coverage
200
Extremely Lightweight Embed
~2.2KB
Guided Emotion and Intonation
Input Streaming for Lower Latency
LLM-based Customizability
Model Tokenizer Type
Non-streaming (CNN-based) tokenizer
Multilingual Support
10+ languages and voice versions available.200+ languages and dialects for Enterprise
Open Source Release Planned
Orpheus Speech Models
Medium (3B)Small (1B)Tiny (400M)Nano (150M)
Pretrained and Finetuned Models
Pretrained modelsFinetuned models
Sample Finetuning Scripts
Sliding Window Detokenizer
Streaming Inference Speed
Faster than playback on A100 40GB for 3B model
Supported Voice Versions
10
Training Data Volume
100k+ hours of speech, billions of text tokens