What do I get back from the Discord file?
A text, one segment per line, with the start and the end time of every segment. The speakers are named on request, and the export comes as TXT, Markdown, SRT, VTT, CSV or DOCX.
Discord call recording to text: upload the recording from Discord and get the text of it with a timecode on every segment, the speakers on request, and TXT, SRT, VTT, CSV or DOCX out. Free $0, 3 files a day.
The file of Discord is all the service needs: our own transcription engine turns it into a text with a timecode on every segment. One file may be up to 30 minutes, and the speakers are told apart on request.
Open the channel and play the voice message, so the client of the device downloads the audio.
The voice message is an OGG container with Opus inside, the format that the client sent.
The waveform of the message shows the shape of the speech, and the length of it is written in the metadata.
The text comes back one segment per line, with the start and the end time of every piece. Fix a name in the editor and export TXT, Markdown, SRT, VTT, CSV or DOCX.
Our own measurement on a short file was 0.84 s. A long recording goes through a queue and is saved to your account, so the text is there when you come back for it.
One segment per line, the way a person wrote it down. It is the format to paste into a ticket, a chat or a document.
The start and the end time of every segment, and the speakers named separately when you ask for it.
SRT and VTT with the time of every line, so the recording can carry its own subtitles.
The transcript is edited through the API and exported as Markdown for a wiki, or as DOCX for a review.
| What | What the platform says |
|---|---|
| Where the file is | A voice message of Discord is one audio attachment: Opus at 48 kHz and 32 kbps in an OGG container. |
| What the file is | The audio of a voice message is Opus at 32 kbps, a compressed file and not a studio recording. |
| What can go wrong | The client sends the audio itself: a voice message is uploaded as a stream of Opus in an OGG container. |
| Length of one file | up to 30 minutes on the free plan, no minute limit on Pro |
| Free plan | $0: 3 files a day, up to 30 minutes per file, no card |
| Pro plan | $10 a month when you pay yearly ($120 a year), or $20 a month |
Say one word of a decision three weeks later and see whether you can find it. In a Discord recording that word lives in a sound file that nobody can skim. Send the file through our own transcription engine and it comes back as text, one segment per line, with the start and the end time of every piece. A search then turns into a quotation with a timestamp.
The audio track is smaller than the video of the same recording.
Check the length before you send it: the free plan takes files of 30 minutes; Pro has no minute limit.
Rename the file before you send it: the name of the meeting travels with it.
Do not guess the language: a wrong setting shows up in the names.
A text, one segment per line, with the start and the end time of every segment. The speakers are named on request, and the export comes as TXT, Markdown, SRT, VTT, CSV or DOCX.
It can be. A voice message is short, and a meeting that was mostly a screen share has little speech in it. The engine returns what is in the file and nothing more.
Not today. The upload widget accepts files only, and no account of ours sits inside Discord. The table above says where the file waits.
That is the rule of Discord and not ours. Take the file as soon as you have it, because a cloud recording can be removed by the platform or by the owner of the account.
Voicemail to text: where iPhone keeps it, and what the engine does with it.
Phone call recording to textPhone call recording to text: where iPhone keeps it, and what the engine does with it.
Podcast episode to textPodcast episode to text: where Podcasts keeps it, and what the engine does with it.
Transcribe any audio fileThe plain version of the service: a file in, a text with the times out.
Speakers by name, a time on every line, export to TXT, SRT, VTT, CSV and Markdown.