n8n
Pull live social and web data into your n8n workflows — and AI agents — with a single verified node. No custom HTTP nodes, no glue code.
Add Social Fetch to n8n and every platform we support becomes a drag-and-drop step: pull profiles, posts, and metrics into a workflow, or attach the node to an AI agent as a live data tool. Each operation maps 1:1 to a public API endpoint, so anything in the API reference works as a no-code step. n8n-nodes-socialfetch is a verified node, so you install it straight from n8n's nodes panel.
What you'll need
- An n8n instance on v1.94.0 or later.
- A Social Fetch API key (starts with
sfk_) — create one in your API Keys dashboard. - Credits on your account for metered routes — see Credits & billing.
Install
Social Fetch is a verified n8n node, so you can add it to your instance straight from the editor. You'll need owner or admin access to install it; once it's in, everyone on the instance can use it.
- Open a workflow and open the nodes panel — click
+or pressN. - Search for Social Fetch and select it from the More from the community section.

- On the node's details page, click Install node.
- Pick an action — for example Get TikTok profile — to add the node to your workflow.


Credentials
Social Fetch authenticates with an API key (it starts with sfk_). Create one in your API Keys dashboard, then connect it from inside the node:
- Open your Social Fetch node and click Set up credential next to the Credential field.

- Paste your key into API Key and click Save.

n8n checks the key against your account as you save it, so an invalid key is flagged right away.
Using the node
Open the Social Fetch node and set up the request:
- Credential — select the credential you just created.
- Resource — pick the platform (e.g. TikTok, Instagram, Web) or Account for whoami/balance.
- Operation — choose the endpoint (e.g. Profile, Search, Markdown).
- Required fields — shown as direct inputs on the node.
- Additional Fields — optional query parameters (sort order, limits, etc.).

The node is marked usable as tool, so you can also attach it to n8n AI agents that need live social or web data.
Pagination and Return All
Only list endpoints that use Social Fetch's cursor query parameter show a Return All toggle. When enabled, the node follows data.page.nextCursor until data.page.hasMore is false. Each page is a separate output item containing the full API JSON body (data, meta, etc.). Use n8n's Item Lists or Code node if you need a single merged array.
List endpoints that return everything in one response (no cursor), or that paginate with other parameters such as page, do not show Return All — you get the first response only unless you pass those parameters under Additional Fields.
Credits
Every request consumes credits from your Social Fetch balance (most endpoints cost 1 credit; some search and media-download options cost more). The credits charged for each call are returned in meta.creditsCharged. See Credits & billing for details.
Example workflow
This example fetches a TikTok creator's recent videos and keeps only the high-performing ones.
- Manual Trigger — so you can run the workflow on demand while testing.
- Social Fetch — set Resource to
TikTokand Operation toProfile videos.- Handle:
n8n - Turn on Return All to page through every video, or leave it off to fetch the first page.
- Under Additional Fields, set Sort By to
Popular.
- Handle:
- Split Out — field to split out:
data.videos(one item per video). - Filter — keep videos above a view threshold:
- Condition:
{{ $json.stats.views }}is greater than100000.
- Condition:
Each Social Fetch item is one API page. With Return All off you get a single page; with it on you get one item per page, each containing data.videos, data.page, and meta. Split Out turns the videos array into individual items so the Filter node can evaluate stats.views on each video.
Example response shape (first video on a page):
{
"data": {
"videos": [
{
"id": "7639528062975053069",
"caption": "…",
"url": "https://www.tiktok.com/@handle/video/7639528062975053069",
"stats": {
"views": 302447,
"likes": 19880,
"comments": 339,
"shares": 2232,
"saves": 389
}
}
],
"page": { "nextCursor": "…", "hasMore": true }
},
"meta": { "creditsCharged": 1, "requestId": "…", "version": "v1" }
}Enrichment pattern
A second common pattern is enrichment: use TikTok → User search (or Search videos) to discover handles, then loop those results back into Profile videos or Profile to pull full details for each match.
Next steps
MCP
Connect Cursor, VS Code, Claude, and other MCP clients to Social Fetch. Sign in with OAuth, call every endpoint as a tool, and let your agent read the docs as it builds.
API reference
Browse Social Fetch endpoints by platform and task, with direct links to generated schema pages and the raw OpenAPI spec