Audio and Video File Support
Custom Vocabulary Support
Emotion Tags
normalslowcryingsleepysighchuckle
Enterprise Grade Security
GDPR compliantISO 27001 certifiedSOC 2 Type II in progress
Guided Emotion and Intonation
High Accuracy Model
Up to 99%
Input Streaming for Lower Latency
LLM-based Customizability
Model Tokenizer Type
Non-streaming (CNN-based) tokenizer
Open Source Release Planned
Orpheus Speech Models
Medium (3B)Small (1B)Tiny (400M)Nano (150M)
Pretrained and Finetuned Models
Pretrained modelsFinetuned models
Real-Time Transcription
<700ms
Sample Finetuning Scripts
Sliding Window Detokenizer
SOC 2 Type II In Progress
Streaming Inference Speed
Faster than playback on A100 40GB for 3B model
Summarization and Sentiment Analysis
Training Data Volume
100k+ hours of speech, billions of text tokens
Transcription Speed
1 hour audio in 10 min
Use-Case Coverage
Creators and MarketersJournalists and EditorsSales and MeetingsDevelopers and BuildersResearchers and AcademicsMedia Monitoring and Broadcasting