3 hours free · No credit card
YouTube, X, Instagram, TikTok. A clean transcript in minutes. With speakers.
Paste a link or POST one. The same speaker-labeled transcript your audience reads on the page is what your code pulls over the wire. Publish-ready in minutes.
Exhibit A · same audio, three minutes apart
Paste a link here
3 hours free · no credit card
right how often does that happen in in one of out of ten deals
how often do you get asked as a seed stage investor to give up your pro rata
you know what's interesting is uh most of the time it's actually yc taking their programmatic
pro rata and they've actually stepped that back so that's actually worked out
really well explain that's the right move they were trying to maintain their 7% indefinitely
How often do you get asked, as a seed-stage investor, to give up your pro-rata?
Most of the time it's actually YC taking their programmatic pro-rata, and they've stepped that back. That's worked out really well.
They were trying to maintain their seven percent indefinitely.
That's right.
Real clip · Garry Tan × Jason Calacanis on This Week in Startups · read the full transcript →
- Studio-grade
- Transcripts
- Speaker-smart
- Labeling
- ~3 min
- Typical turnaround
WHAT PEOPLE SAY
Garry Tan recommends diarize in gbrain
A public shout-out on X. And in his open-source agent-memory project, an instruction that points agents straight at diarize.
From his agent's playbook · gbrain
# Step 1: Get FULL transcript with speaker diarization # WHO said WHAT -- not just a wall of text # Use Diarize.io or equivalent service transcript = diarize(video_url) # speaker-attributed transcriptgbrain · docs/guides/content-media.md →
HOW IT WORKS
Three steps you don't have to take
You paste a link. Everything below happens automatically, in one pass.
Transcribed
The strongest speech model, run end-to-end on the full audio. No chunking artifacts at speaker boundaries.
Diarized
Speaker turns detected at every change of voice and labeled with names you can edit right in your dashboard.
Cleaned
Punctuation, casing, and brand-name corrections, with merge-on-rename. Clean enough to publish as-is.
SEE IT IN ACTION
Watch a transcript come together
Paste a link and the speaker-labeled transcript builds itself: names, timestamps, and clean, publish-ready copy in a couple of minutes.
txt to any video URLFOR HUMANS & AGENTS
Drop it into your pipeline.
A plain REST API that turns any link into a clean, speaker-labeled transcript. One call your pipeline or agent can drive.
Paste. Process. Perfect.
Paste any YouTube, X, Instagram, or TikTok link and get speaker-labeled transcripts in JSON, TXT, SRT, or VTT, ready in minutes.
Built to be called by code.
Submit a URL, poll status, pull the transcript. Bearer auth, no SDK required. Just HTTP.
# Submit a link
curl -X POST https://diarize.io/api/v1/jobs \
-H "Authorization: Bearer $KEY" \
-d '{"sourceUrl":"https://youtu.be/…"}'
# → { "jobId": "job_8a3f…", "status": "CREATED" }
# Poll, then pull the transcript
GET /api/v1/jobs/:id
GET /api/v1/jobs/:id/transcript/json # or srt · vtt · txtUSE CASES
Built for creators and researchers
Whether you're publishing content or analyzing it, get transcripts you can trust.
Give your audience a better experience
Publish studio-quality transcripts anywhere. Caption, embed, and repurpose your videos so they're easier to search and reuse.
Turn videos into searchable knowledge
Pull exact quotes from interviews, lectures, panels, and presentations. Search any conversation by who said what.
QUESTIONS
Frequently asked
Everything you need to know about our transcription service
Still have questions? Contact our support team
Three hours, free
Ready to get started?
Start with 3 hours free. No credit card required.
