Every word, on the record.
A transcript built from the audio, with a speaker on every line and a timestamp on each one — so you can read it, search it, quote it, and check any line against the moment it was said.
Every line
Time-stamped, so you can jump to the audio
Speakers
Labelled by name, not Speaker 1 and Speaker 2
Search
Full-text search inside the episode and the library
13
Languages you can read the transcript in
A real transcript from 99% Invisible
Speaker labels, timestamps and search, on one episode. The highlighted line is a search hit.
How it works
From an episode to Transcript, in three steps.
- 1
Process an episode
Open a podcast you follow, or add an episode or a YouTube link. Podwise transcribes the audio.
- 2
Read with the audio
Every line carries its timestamp and speaker. Search a word, and the matching lines light up.
- 3
Quote or export it
Copy the lines you need, export SRT or Markdown, or fetch the transcript from the Open API.
What you get
Everything comes from the episode you already processed.
- Speaker labelsHost and guests are named, so a quote is attributable and a conversation is easy to follow.
- A timestamp on every lineEvery sentence points back to the second it was spoken, so you can verify a name or a number.
- Search and highlightsSearch inside one episode or across your whole library, then jump to the line and the audio.
- TranslationsRead the transcript in any of the 13 supported languages.
- ExportsSRT for subtitles, Markdown for notes, and a transcript endpoint on the Open API.
- Accuracy hintsWhen an episode goes in through the API, you can pass keywords and speaker names so unusual names are recognised.
More features
Every one works on the same episode, so nothing has to be processed twice.
Mind map
The whole episode as one map of its chapters and ideas.
Summaries
Overview, chapters, key Q&A and highlights for every episode.
Ask
Ask one episode, or your whole library, a real question.
Translation
Read any episode in 13 languages, side by side.
Transcript questions
How accurate is the transcript?
It is generated from the audio by AI, and it is clean enough to read, search and quote. Names, jargon and people talking over each other are where mistakes appear; the timestamp on every line is there so you can check the audio.
Can I improve accuracy for one episode?
Yes, on the API: pass the speaker names and any keywords or proper nouns when the episode is created, and they are used to improve recognition. The clearer the audio, the better the result.
Can I search inside an episode?
Yes. Search finds the word or phrase inside the transcript, and the result takes you to the line and its place in the audio. Search also runs across the episodes in your library.
Can I export the transcript?
Yes: SRT for subtitles, Markdown for notes, and the transcript endpoint on the Open API for anything else.
Every word, searchable.
Start with Podwise for free and open the transcript on any processed episode.