How to read on-screen text at a given second

Use POST /v1/watch_video with a window of one second. That second is read at full resolution on our side, and what comes back is what is actually there rather than what a small picture suggests. Add a region when the caption sits in one corner.

Last updated 31 August 2026

The call

Copy this as it stands. “What does the caption at 0:37 say, and how is it set?” is the question it answers.

Ask an agenttext

Read the text on screen at 0:37 in this video and describe how it is set.

cURLbash

curl -X POST https://api.tryplayhead.com/v1/watch_video \
  -H "Authorization: Bearer $PLAYHEAD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "…", "window": {"from": 37, "to": 38}, "question": "What does the caption say?"}'

Watch out for

Everything the API can refuse, and what to send instead, is on the errors page.