TikTok or Instagram post URL to media details: accepted URL shapes

To get details for one TikTok or Instagram post on Sume, send its public URL to POST /v1/scrapecreators/tiktok_media or POST /v1/scrapecreators/instagram_media. The URL must be https, on tiktok.com or instagram.com, and point at a single post; a profile, story or login link is rejected with 400 scrapecreators_invalid_input.

The answer is a media candidate: metadata and links, not a copy hosted by Sume. Copying the bytes is the separate media import step.

URL shapes that pass

Instagram paths must start with /p/, /reel/ or /reels/. TikTok paths must be @user/video/<id>, @user/photo/<id> or the short /t/<code> form, and the short-link hosts vm.tiktok.com and vt.tiktok.com are accepted. Query strings and hash fragments are stripped before the lookup, so tracking parameters on a pasted link are harmless.

instagram_media also takes a shortcode instead of a url, but never both at once. tiktok_media requires url. Both accept region, a two-letter code.

Call it

Pass the pasted link as is. If you hold only a handle, use the profile and feed routes first and take the post URLs from there.

curl -sS https://api.sume.com/v1/scrapecreators/instagram_media \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://www.instagram.com/reel/EXAMPLECODE/"}'

From candidate to something you can generate with

A candidate lets you decide whether a post is worth keeping. To use it as an input for a video model or for analysis, import it: POST /v1/media-imports takes a public TikTok or Instagram video URL, mirrors it to media.sume.com and records an asset, at a fixed estimate of $0.15 per accepted import. YouTube and other hosts are rejected there with unsupported_platform.

Reading is free in v1 and importing is not, so the cheap pattern is to read many candidates, filter on the metadata you got back, and import only the few you will actually use. The import cost breakdown has the per-batch arithmetic.