Retrieve a capture

{captureId} is the id from the create response. It is a UUID; anything else is a validation failure rather than a 404. Authenticate with the same X-Api-Key header, and note that a key only sees captures made by its own account.

curl https://api.websitescreenshotworld.com/instant-snap/v1/captures/0f9d1a6c-3b2e-4f18-8c6a-1d4b7e9a2c05 \
  -H "X-Api-Key: $WSW_API_KEY"

The response is the same envelope documented on Create a capture, with the same fields and the same meanings. There is no separate read schema to learn.

download_url is signed and valid for 60 minutes. It is generated when you read the capture, so a link that has gone stale is not a problem — read the capture again and you get a new one. expires_at does not move: it is when the file itself is deleted, 24 hours after the capture.

The one capture you cannot read back is a by_format one. Those return the bytes and store nothing, so their id has no record behind it and this endpoint answers capture_not_found.

An async capture reads queued until a browser takes it, then running, then succeeded or failed like any capture. Read it every few seconds; the client libraries' wait does this for you. A capture that never got a browser within 15 minutes reads failed with error_code: capture_expired and was not charged.

Once the media has been deleted the record survives, so a read still answers 200 — withstatus: expired, a null download_url and a null expires_at:

{
  "success": true,
  "id": "0f9d1a6c-3b2e-4f18-8c6a-1d4b7e9a2c05",
  "status": "expired",
  "response_type": "json",
  "media_category": "image",
  "format": "png",
  "content_type": "image/png",
  "size_bytes": 184203,
  "width": 1920,
  "height": 1080,
  "duration_ms": 3184,
  "page_http_status": 200,
  "captured_at": "2026-08-31T09:14:22.000Z",
  "final_url": null,
  "media_sha256": "9f86d081884c7d659a2feaa0c55ad015a3bf4f1b2b0b822cd15d6c15b0f00a08",
  "download_url": null,
  "expires_at": null,
  "page_source": null,
  "s3_delivery": null,
  "callback_url": null,
  "error_code": null,
  "spec_version": "v1"
}

If you need a capture for longer than a day, copy it into your own storage while the link is live — or have the API write it there for you with S3 delivery, which is not subject to our retention at all.

{
  "success": false,
  "error_code": "capture_not_found",
  "error_message": "\"id\" was not found"
}

• The id does not exist.
• The id belongs to another account. The API answers 404 rather than 403 on purpose — the alternative would let anyone confirm which capture ids exist.
• The capture was made with response_type: by_format and therefore never stored.