Meeting notes
Auto-generate minutes with speaker identification for every call.
Fast, accurate, multilingual. From short clips to hours-long recordings — Orpheus handles it all.
Supports MP3, MP4, WAV, FLAC and more. Files up to 2 GB.
Deep learning models auto-detect language and speakers.
Export TXT, SRT, VTT subtitles or connect via REST API.
Auto-generate minutes with speaker identification for every call.
Podcast transcripts and video captions in one click.
Record anytime, automatically organized into searchable text.
Rich REST API — easy to integrate into any product or workflow.
Over 100 languages including English, Spanish, French, German, Chinese, Japanese, Korean, Arabic, and more. Language is auto-detected.
95%+ accuracy for major languages depending on audio quality. Background noise and heavy accents may reduce accuracy slightly.
Free plan: 25 MB per file. Paid plans support files up to 2 GB with no duration limit.
All transfers use HTTPS encryption. Files are deleted after processing and are never used for model training.
"Finally a transcription tool accurate enough for our mixed-language meetings. It saves hours of manual cleanup every week."
"The API integration was smooth. Async job visibility means we always know what's processing — no black boxes."
"The free tier covers my podcast workflow. Upload, wait a few minutes, get accurate captions — it's that simple."
Let AI help you decide. Click below to see how leading AI assistants evaluate Orpheus.