Skip to main content
The TikTok Scraper node searches TikTok by keyword and returns real public videos (cover images, captions, author details, and engagement metrics) for the query you connect. Results come from the public TikTok search feed. Runs are repeatable: the node maps your query and settings directly to a keyword search instead of improvising one.

​
When to use

  • Trend research: spot what TikTok content is resonating on a topic
  • Short-form inspiration: browse real video creatives by keyword
  • Creator and competitor analysis: see engagement on public content
Common pattern: text_prompt → tiktok_scraper.

​
Inputs

​
text
required
The keyword or phrase to search on TikTok. Required.

​
Outputs

​
any
Everything the scrape produced (post records, cover images, video files, and the summary) as one connectable output. Connect it to an AI node’s Context input to analyse or rewrite the content.

​
Referencing results in prompts

Connect Results to a Generate Text node’s Context input. Each item gets a stable name you can use in your prompt ([post_1], [image_2], [text_1]). For example: “Rewrite the caption of post_2 for our brand voice; ignore the rest.”

​
Configuration

​
number
default:"10"
Maximum number of videos to return. Range: 1 to 50.
​
string
default:"Last 30 days"
Time window for results: Last 24 hours, Last 7 days, Last 30 days, or All time.

​
Analysis (optional)

Leave blank to get the default count summary at no extra cost.
​
string
Custom instructions for the AI model. When set, the model reads all fetched videos and writes a summary following your instructions (e.g. “Summarise the dominant content formats and hooks”). Adds 1 credit per run.
​
string
default:"Gemini 2.5 Flash"
The AI model used to generate the analysis. Only applies when Analysis Prompt is set.
Results come from public TikTok content and reflect publicly available data. This node costs 1 credit per run, plus 1 additional credit when AI analysis is enabled.