Build with data
YouTube transcripts.
Ready for your workflow.
Turn available video captions into readable text and timed segments. Read one video or a batch, inspect each result and pay only for successful items.
Public data through the APINo card required
Illustrative response fields. See the recorded batch example below for live evidence.
- Readable transcript text
- Timed segments where available
- Per-item failure accounting
The workflow
From video IDs to usable text.
- 01
Choose the videos
Send public YouTube video IDs in a batch request.
- 02
Read each result
Check each item’s success, transcript and timed segments.
- 03
Handle missing captions
Failed items remain visible and cost zero credits. Retry only the items you still need.
What you get
One request. Individual results.
Batch reads preserve the result and charge for each video.
- 01Text
- Readable caption text when the source supplies it.
- 02Timing
- Segment timestamps for navigation or citation.
- 03Billing
- Existing per-video prices, with unsuccessful items uncharged.
Plan your integration
Know the scope.
Sources and availability
Public videos with captions available to the current source. Batch requests preserve video IDs and per-item status.
What to keep in mind
This retrieves available captions; it does not generate speech recognition for every video. Removed, private, restricted or uncaptioned videos may fail. A small timing sample is not a latency guarantee.
Questions, answered
About youtube transcripts.
Can every video be transcribed?
Only captions available from the source are returned. A private, restricted or uncaptioned video can fail without a credit charge.
How much does a batch cost?
The sum of the existing per-item prices. Failed items cost zero credits; inspect each item rather than only the outer response.
Do my credits expire?
Purchased pack credits never expire. Monthly subscription and Free allowances reset.
Can an agent retrieve the full response?
The hosted MCP connector supports get_result for stored results at zero retrieval credits. Use the returned retrieval reference.
Two videos, two readable transcripts
Recorded on 21 September 2026: both videos returned text and timed segments. Three fresh batches took 2.193, 1.975 and 1.952 seconds. This is a small functional sample, not a speed guarantee.
curl -X POST 'https://www.monocrawl.com/v1/youtube/transcripts' \
-H "Authorization: Bearer $MONOCRAWL_API_KEY" \
-H 'Content-Type: application/json' \
-d '{"ids":["dQw4w9WgXcQ","jNQXAC9IVRw"]}'Recorded result summary
{
"success": true,
"count": 2,
"succeeded": 2,
"failed": 0,
"credits_used": 4,
"billing_complete": true
}Current base price: 4 credits, allocating $0.01180 from the $59 / 20,000-credit pack, before tax. $0.00590 per successful transcript.
This is the recorded response summary. Each item contains its own status and data. Missing captions do not become an invented transcript. Pack cost allocation assumes all purchased credits are used; the checkout payment is $59.
Start with a real request
Try the data. Then build the workflow.
1000 free credits. No card required.