Loading...
A short, honest log of what we shipped.
The Upload button on the homepage is now a real drop target. Drag a file from your desktop, drop it onto the button, and you go straight into the upload flow.
Click still works exactly the same as before. The button just got a second job.
If you uploaded a recording in the last 90 days, you will see a player at the bottom of the transcript page when you open it. Press play and the active block lights up as it speaks; you can scrub to any moment, or click a block's timestamp to jump straight to that line.
Space toggles play and pause. Left and right arrows skip ten seconds. The speed control remembers what you last set. None of these shortcuts steal focus from the editor — you can listen, edit, and listen again without juggling.
If you stopped using TranscribeCat for more than 90 days, your audio has been deleted automatically. The transcript still works; the player will just be hidden, and we tell you why.
You no longer have to drop a transcript into Word to fix a misheard name. Open any transcript and click into a block — type your correction, click out, and it saves. Cmd+Enter saves; Escape reverts.
Speaker labels are clickable too. Rename 'Speaker 1' to 'Marcos' once, and every block by that speaker updates everywhere — including in the file you download next.
Your downloads (Text, Subtitles, Word) now reflect your edits. The original AI transcript is preserved in our database, untouched, so we can always go back to it.
If you uploaded a recording over about three hours and it got stuck in 'Processing' without finishing — that won't happen now.
Two things were wrong: our worker had a one-hour ceiling that was shorter than long files needed, and our chunk size was right on the edge of the transcription API's own limit. We fixed both, and we also added a safety net that detects when a job runs out of time and refunds the charge automatically rather than leaving it stuck.
If you had a transcript stranded by this, it has already been refunded.
We tightened the visual design across the landing page, the dashboard, and the transcript view. One typeface ramp, one accent color, one button style. Loading spinners now use a small soundwave instead of the generic spinner.
Nothing about the workflow changed. If a part of the site looks different, that is why.
Previously, we deleted your uploaded audio the moment your transcript was ready. We now keep it for 90 days from your last sign-in. While you are using TranscribeCat regularly, your audio stays around. If you stop using your account, it gets deleted automatically.
Why: it lets us replay and verify your transcript when you want to compare it to the source, and it sets us up for a player on the transcript page (coming soon).
You can still delete your audio at any time from your profile, or delete a single transcription, or delete the whole account. The Privacy and Trust pages explain the new policy in plain terms.
If you are doing thesis interviews, cutting podcast episodes, sitting through lectures, or running team meetings, there is now a page that explains how TranscribeCat fits your specific work.
You can find them at /for/researchers, /for/podcasters, /for/students, and /for/businesses, or via the footer.
We added an /about page covering who TranscribeCat is for, who builds it, and why we charge per hour instead of per month.
It's short. We will keep it that way.
Have a question about something here, or something you wish we'd ship? Open the in-app inbox after signing in and tell us.