Business Terminology Support
Context-Aware Translation
Emotion Tags
normalslowcryingsleepysighchuckle
Enterprise-Grade Translation
Guided Emotion and Intonation
Input Streaming for Lower Latency
Language/Country Support
50+ languages
LLM-based Customizability
Model Tokenizer Type
Non-streaming (CNN-based) tokenizer
Multilingual Conversation Training
No Download Required (Web)
Open Source Release Planned
Orpheus Speech Models
Medium (3B)Small (1B)Tiny (400M)Nano (150M)
Pretrained and Finetuned Models
Pretrained modelsFinetuned models
Professional-Grade Voice Translation
Real-time Voice Translation
Sample Finetuning Scripts
Simultaneous Interpretation
Sliding Window Detokenizer
Streaming Inference Speed
Faster than playback on A100 40GB for 3B model
Training Data Volume
100k+ hours of speech, billions of text tokens
WebRTC Video Conferencing