Skip to content

Help

Support and documentation

Short guides for the things people ask about most, plus a direct line when they don't cover it.

Getting your first transcript

Upload a file, watch it process, and read the result. Takes about three minutes end to end.

Fixing speaker labels

Rename, merge and split speakers, and reassign a single line the model attributed wrongly.

Choosing an export format

Which of the eight formats to use, and what each of the seven document templates is for.

Working with long recordings

Chapters, search within a transcript, and keeping a three-hour file navigable.

Importing from a podcast feed

How feed and episode links resolve, and what to do when a platform blocks direct access.

Translating a transcript

Producing a second language version while keeping timings and speaker attribution.

Contact us

Questions about a transcript, a plan, or whether Clea fits what you are doing. We read everything and reply within one working day.

hello@clea.app
Not yet available

API

A REST API for submitting media and retrieving transcripts is planned for the Business plan. It is not built yet, and we would rather say so than list endpoints that do not exist.

See what Business includes today

Something not working?

If a transcription failed, include the project name in your email — it lets us look up exactly what happened without asking you to reproduce it.

FAQ

Questions people ask first

Short answers to what people ask before their first upload: which file formats work, how speaker detection behaves, which languages are supported, what you can export, and what happens to a recording once it has been processed. If something here is missing, the help centre goes further.

MP3, WAV, M4A, AAC, OGG, FLAC, MP4, MOV, WebM, MKV and most other common formats. If your file plays in a browser or a media player, Clea can almost certainly read it.

Yes, as long as you upload the file. Clea does not download from those platforms — doing so generally conflicts with their terms of service — but if you have the rights to a video, save it to your device and upload it. The transcription is identical.

No. Upload whatever you have — MP3, WAV, M4A, MP4, MOV, WebM and most other formats work as they are. For video, Clea reads the audio track and ignores the picture.

Yes. Speaker diarization runs on every transcription, splitting the conversation into labelled turns. Clea does not know anyone's real name, so speakers start as Speaker 1, Speaker 2 and so on.

Yes, and renaming propagates everywhere at once — the transcript, the summary, the AI insights and every export.

Both, plus TXT, SRT, VTT, CSV, Markdown and JSON. PDF and Word exports use one of seven document templates and let you choose whether to include timestamps, speaker names, the summary, chapters and page numbers.

Yes. After transcription you can translate into another language while keeping the original timings and speaker attribution, then export the translated version.

Accuracy depends heavily on the recording. Clear speech on a decent microphone is typically very accurate; heavy accents, crosstalk, background noise and distant microphones all reduce it. Clea shows per-segment confidence so you can see where to check.

Usually a fraction of the recording's length — a one-hour file typically finishes in a few minutes. You can close the tab while it runs; processing continues on our side.

Uploads travel over TLS and are stored in private buckets that are not publicly listable. Media is only ever read through short-lived signed URLs. Your projects are readable only by your account, enforced at the database level rather than only in the application.

Files stay until you delete them. Deleting a project removes its media, transcript, insights and generated documents. You can also set a project to delete its source media automatically once transcription completes.

Yes. Search runs across every transcript in your account and returns the matching sentence with its speaker and timestamp. Clicking a result opens the recording at that exact moment.