Documentation Index
Fetch the complete documentation index at: https://docs.vidjutsu.ai/llms.txt
Use this file to discover all available pages before exploring further.
The /v1/scrape/* endpoints are thin, typed wrappers over the public-data surfaces of TikTok, Instagram, X, YouTube, and the Meta / Google / LinkedIn / Reddit ad libraries.
You don’t sign up for a scraper. VidJutsu holds the upstream master credentials and meters the calls through your existing API key. One key, one bill, one ToS.
Why a wrap
Building a creator-research, ad-research, or content-intelligence product normally means juggling a separate ScrapeCreators (or equivalent) account, a separate bill, a separate set of rate limits, and a separate ToS surface. The scrape primitives collapse that into your VidJutsu key — same auth, same balance, same operational logging — and stay typed in our OpenAPI spec so they generate clean SDK methods.
Most endpoints accept an optional download_media: boolean (default false).
false (default) — Response contains raw source CDN URLs (TikTok / IG / X / YouTube / ad-library host URLs). Cheapest. Caveat: these URLs are often gatekept and not fetchable downstream from VidJutsu Watch / Transcribe / external tools like Gemini.true — VidJutsu fetches every media URL in the response, stages it to our public CDN, and replaces the source URLs with VidJutsu URLs in the response. Use this when you need to pipe the media into a downstream pipeline (Watch, Transcribe, an LLM video tool, your own CDN, etc.).
This matches the same staging pattern used by POST /v1/upload/url.
Op-counting
| Cost | Trigger |
|---|
| 1 credit | Base cost per call. Always charged. |
| +1 credit per staged media file | Only when download_media: true and the response contains media URLs that VidJutsu pulled onto its CDN. |
scrapeInstagramUserPosts call that returns 12 posts with download_media: true and stages 12 videos costs 1 + 12 = 13 credits. The same call with download_media: false costs 1 credit.
Auth & rate limits
- Auth: Paid (subscription required, same as Watch / Extract / Transcribe).
- Rate group:
scrape — 500 calls per day, shared across all 24 endpoints. Separate bucket from intelligence (Watch / Extract / Transcribe / Check) and compliance.
- Check remaining capacity with
vidjutsu usage or GET /v1/usage.
The 24 endpoints
All are POST and accept a JSON body. Full request/response schemas in the OpenAPI spec.
TikTok
| Endpoint | Body | What it returns |
|---|
POST /v1/scrape/tiktok/profile | handle, trim?, download_media? | Profile metadata for a single TikTok user |
POST /v1/scrape/tiktok/profile/videos | handle, cursor?, download_media? | Paginated list of a user’s videos |
POST /v1/scrape/tiktok/video | url, trim?, download_media? | Full metadata for a single video |
POST /v1/scrape/tiktok/video/transcript | url | ASR transcript for a single video |
POST /v1/scrape/tiktok/video/comments | url, cursor? | Paginated comments on a video |
POST /v1/scrape/tiktok/search/users | query, cursor? | User-search results for a query |
POST /v1/scrape/tiktok/trending | country?, download_media? | Trending feed for a country |
Instagram
| Endpoint | Body | What it returns |
|---|
POST /v1/scrape/instagram/profile | handle, download_media? | Profile metadata for an IG user |
POST /v1/scrape/instagram/user/posts | handle, cursor?, download_media? | Paginated posts for a user |
POST /v1/scrape/instagram/post | url, download_media? | Full metadata for a single post or reel |
POST /v1/scrape/instagram/post/comments | url, cursor? | Paginated comments on a post |
POST /v1/scrape/instagram/user/reels | handle, cursor?, download_media? | Paginated reels for a user |
| Endpoint | Body | What it returns |
|---|
POST /v1/scrape/twitter/profile | handle, download_media? | Profile metadata for an X user |
POST /v1/scrape/twitter/user-tweets | handle, cursor?, trim?, download_media? | Paginated tweets for a user |
POST /v1/scrape/twitter/tweet | url, trim?, download_media? | Full metadata for a single tweet |
POST /v1/scrape/twitter/tweet/transcript | url | Transcript for a tweet that contains a video |
YouTube
| Endpoint | Body | What it returns |
|---|
POST /v1/scrape/youtube/channel | handle? or channel_id?, download_media? | Channel metadata |
POST /v1/scrape/youtube/channel-videos | handle? or channel_id?, cursor?, download_media? | Paginated videos for a channel |
POST /v1/scrape/youtube/video | url, download_media? | Full metadata for a single video |
POST /v1/scrape/youtube/video/comments | url, cursor? | Paginated comments on a video |
Ad libraries
| Endpoint | Body | What it returns |
|---|
POST /v1/scrape/ads/meta | query, country?, download_media? | Active ads from the Meta Ad Library for a keyword or page name |
POST /v1/scrape/ads/google | company, download_media? | Active ads from Google’s Ad Transparency Center for an advertiser |
POST /v1/scrape/ads/linkedin | query, download_media? | Active ads from the LinkedIn Ad Library for a keyword |
POST /v1/scrape/ads/reddit | query, download_media? | Active ads from the Reddit Ad Library for a keyword |
CLI
Every endpoint is exposed as a hyphenated vidjutsu scrape <subcommand>:
vidjutsu scrape tiktok-profile --handle mrbeast
vidjutsu scrape instagram-user-posts --handle nike --download-media
vidjutsu scrape meta-ads --query "ai video" --country us
Example
curl -X POST https://api.vidjutsu.ai/v1/scrape/instagram/user/reels \
-H "Authorization: Bearer $VIDJUTSU_API_KEY" \
-H "Content-Type: application/json" \
-d '{ "handle": "nike", "download_media": true }'
video_url and thumbnail_url rewritten to a stable VidJutsu CDN URL you can hand to /v1/watch, /v1/transcribe, or any third-party tool.
