FeedMCP

Data APIs / Douyin

Douyin related keywords API

The words most related to a keyword on Douyin for a week: rank, search index and change in relevance.

Item Value
Endpoint GET https://api.feedmcp.com/v1/douyin/index/keywords/related
ID douyin.index.keywords.related
Price $0.002 / call
Returns A page of Trend objects
Auth Authorization: Bearer $FEEDMCP_API_KEY

Ocean Engine computes relations by week (Monday to Sunday), so end_date snaps to the Sunday before it.

Parameters

Name Type Required Description
keyword string yes Keyword
end_date string no Which week: the week ending on the last Sunday before this date; default last week

Every endpoint also takes view=full to include the platform's raw data as raw, at the same price.

Returned fields

data.items is a list of Trend objects, with the same fields on every platform:

Field Type Description
rank number Rank; 0 means pinned
platform string Platform
keyword string Term
heat number | null Heat (each platform counts its own way)
label string | null Label, such as pinned or hot
url string | null Search link on the platform

When you are charged

$0.002 per call that returns the list. An empty list (the platform did not answer), platform errors and timeouts are free; just retry.

Each answer carries meta.billing: what this call cost and the balance left.

Request

cURL

curl "https://api.feedmcp.com/v1/douyin/index/keywords/related?keyword=%E9%9C%B2%E8%90%A5" \
  -H "Authorization: Bearer $FEEDMCP_API_KEY"

Python

import os, requests

r = requests.get(
    "https://api.feedmcp.com/v1/douyin/index/keywords/related",
    params={"keyword": "露营"},
    headers={"Authorization": f"Bearer {os.environ['FEEDMCP_API_KEY']}"},
    timeout=30,
)
print(r.json())

JavaScript

const res = await fetch("https://api.feedmcp.com/v1/douyin/index/keywords/related?keyword=%E9%9C%B2%E8%90%A5", {
  headers: { Authorization: `Bearer ${process.env.FEEDMCP_API_KEY}` },
});
console.log(await res.json());

MCP

From an AI with FeedMCP installed, the call tool:

{
  "endpoint": "douyin.index.keywords.related",
  "params": {
    "keyword": "露营"
  },
  "max_price": "0.002"
}

Response

The data below is made up; the shape is the real one.

{
  "code": "ok",
  "request_id": "req_7Qm2kXb9TzLp",
  "data": {
    "items": [
      {
        "rank": 1,
        "platform": "douyin",
        "keyword": "户外露营",
        "heat": 229638,
        "label": "关联度变化 -0.88%",
        "url": "https://www.douyin.com/search/%E6%88%B7%E5%A4%96%E9%9C%B2%E8%90%A5"
      },
      {
        "rank": 2,
        "platform": "douyin",
        "keyword": "露营装备",
        "heat": 180000,
        "label": "关联度变化 1.2%",
        "url": "https://www.douyin.com/search/%E9%9C%B2%E8%90%A5%E8%A3%85%E5%A4%87"
      }
    ]
  },
  "page": {
    "next_cursor": null,
    "has_more": false
  },
  "warnings": [],
  "meta": {
    "endpoint": "douyin.index.keywords.related",
    "platform": "douyin",
    "fetched_at": "2026-11-19T08:00:00.000Z",
    "source": "FeedMCP",
    "catalog_revision": "d09a736bd39e8b3f",
    "billing": {
      "charged": true,
      "amount": "0.002",
      "currency": "USD",
      "balance": "9.98"
    }
  }
}

Errors are described in Errors: each says whether you were charged (charged) and whether to retry (retry).

Other Douyin endpoints

Questions

$0.002 per call that returns the list. An empty list (the platform did not answer), platform errors and timeouts are free; just retry.

Do I need a platform account?

No. Only public data is read.

A video ID is the number after /video/; an account is the sec_user_id after /user/, not the Douyin handle.

How do I get more items?

When the answer has page.next_cursor, pass it back as cursor for the next page. Each page is billed on its own.

How do I use it from an AI?

Install the MCP server and ask; your AI calls douyin.index.keywords.related.