A video and its author

The video command is the most reliable in the tool. It reads the JSON the video page already ships, so it needs no signing and answers from anywhere.

tt video https://www.tiktok.com/@tiktok/video/7106594312292453675 -o json

The record carries the caption, the create time, the author, the sound, the hashtags, the dimensions, the playable urls, and every counter. The author block includes a author_sec_uid, the opaque id the posts endpoint needs.

From one video to the author's feed

posts takes a handle or a secUid. Given a handle it resolves the secUid from the profile page first, then pages the feed:

tt posts @tiktok -n 60 -o jsonl

If you already have a secUid from a video record, pass it directly to skip the lookup:

tt posts MS4wLjABAAAAv7iSuuXDJGDvJkmH_vz1qkDZYo1apxgzaxdBSeIuPiM -n 60

posts rides the signed API plane, so from a datacenter IP it may exit 4. See troubleshooting.

Just the links

Every record carries a url, so -o url gives a clean link stream:

tt posts @tiktok -n 100 -o url > links.txt

Picking columns

--fields selects and orders columns for the table and the delimited formats:

tt posts @tiktok -o csv --fields id,create_time,play_count,digg_count