Upload a recording.
Get who said what, and when.
Orea transcribes Papiamentu, separates the speakers and times every line. Low-confidence words are marked so you know where to look first.
New upload
How it will appear in your list. Defaults to the file name.
Orea is a Papiamentu model; other languages open as they are measured.
Sets the language model that helps with wording.
If you know how many, say so: it keeps two similar voices apart. You can change it afterwards.
A translation, when available, appears under each sentence.
You get an email when the transcript is ready. Every upload and every correction you make may be used to improve Orea; only upload recordings you have the right to process. Terms · Privacy
Your recordings
| Title | Length | Speakers | Status | |
|---|---|---|---|---|
| Loading… | ||||
Jobs run one at a time on one GPU, taking turns between users. A dense hour of meeting audio is roughly 20 minutes of processing; you get an email when it is done.