TikTok trending video transcripts via API: what summary_mode returns
Can you get a transcript with Sume's trending video search? summary_mode transcript returns metadata plus an unsupported warning. What to do for the words.

No. Sume's trending video search returns metadata, and summary_mode: transcript currently returns that metadata plus an unsupported warning rather than the spoken words. The download and download_limit fields behave the same way: values above zero return an unsupported warning because mirroring is not built.
If you need the words of a trending clip, you need the clip in your own workspace first, and then a transcript call on it.
What does trending search return?
Ranked public TikTok video metadata: a canonical watch url, an optional cover_url thumbnail (not a downloadable video), description, created_at, region, author.handle, metrics, relevance and scores. Each accepted call costs $0.10, and summary_mode: metadata is included in that price. Confirm live in GET /v1/catalog.
curl -X POST https://api.sume.com/v1/trending-videos/search \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query":"skincare routine","window":"this-week",
"limit":10,"summary_mode":"metadata"}'Which summary_mode values exist?
The field is none (default), metadata or transcript.
| Value | Result |
|---|---|
none | Videos and metrics only |
metadata | Lightweight summary from metadata, same price |
transcript | Metadata plus an unsupported warning |
How do I get the words of a clip I found?
Sume does not hand you the raw TikTok CDN file, and the docs say not to expect downloadable sources from this endpoint. If you have the right to use the clip, bring it into your workspace as a media.sume.com artifact (for example from a media import of a file you own), then run video inspect with transcribe: true at about $0.01 per audio minute, or reference ingest with speech.allow_billed_stt.
Importing someone else's video is a rights question that Sume's tooling does not answer for you. Read the creator's and the platform's terms.
How do I plan research on a trend without transcripts?
Use the metadata for what it is good at: which hashtags and captions recur, which regions a trend sits in (browse on dev fans out across US, GB and KR when no region is given), and how metrics compare inside a window such as this-week or this-month. Production generic queries default to this-month, and known SaaS profiles default to last-3-months.
Then pick two or three clips, get lawful copies into your workspace, and run the analysis surfaces on those only. That keeps the paid steps small: one $0.10 search call, then per-clip inspection.
Is the browse feed live in production?
The no-query browse grid, 1 to 100 limit and quality filters are on for dev. Production keeps the legacy contract (query required, limit 1 to 50, default 10) until SUME_COM_TRENDING_VIDEOS_REBUILD_ENABLED is exactly 1 or true. Platform support is TikTok only.
Sources
Related posts
More in Media tools
- Trim an AI clip first, or use source_in on the timeline?
For a shot that is too long, set source_in and duration on the Timeline 1.0 slot. Use video_trim ($0.02 per job) only when another step needs a new MP4 file.
- Put a product shot above a UGC creator clip: compose stack
Timeline compose stacks one Sume-hosted still over one video into a single MP4. Set ratio for the still's share. The public rate is $0.02 per job.
- Check a video filter program for free before you encode it
POST /v1/video-filter/check runs the same schema and allowlist as the encode and returns diagnostics without creating a job. Encoding is $0.02 a job.
- Video frames vs video inspect: max_edge 16 vs 64 and the 768 default
Sume video-frames max_edge is 16 to 2160 and keeps source size if omitted. Video-inspect max_edge is 64 to 2160 and defaults to 768.
Written by Sume