How is the sponsor read found?
Export the transcript and search the name of the sponsor in the text: every line carries the minute it starts at, so the read is found by the word and the minute tells you where to listen again.
Podcasts
Transcribe a podcast episode to text
Podcast to text: the minute of every quote, the chapters for the show notes and the sponsor read to check, with the start and the end time on every line.
3 files a day up to 30 minutes, free99+ languagesSpeaker separation and timestamps
In short
Upload the audio of the episode and the transcript comes back with the start and the end time of every line: the minute of a quote, the chapter for the show notes, the sponsor read. One file is up to 30 minutes, and the free plan is $0.
How to transcribe a podcast episode
- 1
Take the episode file as it is published
The file a listener downloads is the one to transcribe. Keep it untouched, without trimming the intro or the outro: the timecodes of the transcript are counted from the beginning of that file.
- 2
Upload the episode and split the voices
One file is up to 30 minutes. Choose the language or leave the detection on, and turn on the split by speaker when the host and the guest sit on different microphones.
- 3
Find the minute of the quote
Every line shows the time it starts and ends. SRT, VTT, CSV and the table of the Markdown file carry those numbers, while TXT and DOCX on Pro carry the words for the show notes.
- 4
Write the chapter list
The chapters of an episode are written from the text and the minutes: open the exported file, note the minute of every change of subject, and paste the list into the description of the show.
What you get
- TXT
Plain text for an email, a note or a chat
- SRT
Subtitles for video editors and platforms
- VTT
Subtitles for a browser and a website
- CSV
A table: time, speaker, text
- DOCX
A Word document with speakers and times
- Markdown
Text with headings for notes
What the transcript of an episode holds
| What | Value |
|---|---|
| One episode | up to 30 minutes on the free plan, no minute limit on Pro |
| A longer episode | split into parts before the upload; the service does not cut the file for you |
| The two hosts | each voice gets a label of its own when the split by speaker is on |
| Free plan | $0 with 3 files a day, up to 30 minutes per file, no card |
The case
Why the editor needs the minute, not the audio
An episode is talked for an hour and then heard for an hour by every listener. The people who run a show need it the other way round: the exact sentence, with the minute it was said in, so a clip can be cut and a chapter can be written without scrubbing through the audio.
Tips
Number the parts of a long episode in the name of the file.
Split by speaker when the show records on two microphones.
Export SRT when the episode has to become a subtitled video.
Write the chapter list from the minutes, not from memory.
Questions this page answers
Can the host and the guest be separated?
Yes, switch on the split by speaker. A host and a guest then get a label each, and a chapter can be attributed to the person who actually said it.
Do you publish an accuracy percentage?
We do not, and we have not measured one on our own recordings: a number without the test set behind it would mean nothing. What we publish is the limit of the file, the limits of the plan and the formats of the export.
What if the episode is longer than 30 minutes?
One file is up to 30 minutes, on every plan, and the service does not cut the file for you. A longer episode is split into parts before the upload; every part is a transcript of its own, and its timecodes start again at zero.
Reading: 6 minutesUpdated 
Transcribsi editorial team
Drop a recording — the text arrives in a minute
Speakers by name, a time on every line, export to TXT, SRT, VTT, CSV and Markdown.
3 files a day up to 30 minutes, free99+ languagesSpeaker separation and timestamps