Toutiao Scraper

Collect public article and video search results from Toutiao (今日头条) by keyword. The Actor turns Toutiao's search pages into structured records for Chinese media monitoring, topic research, brand tracking, and data pipelines.

Use Chinese-language keywords for the strongest coverage. Choose article results, video results, or both, and set a hard output limit that is also the maximum number of charged items.

What does Toutiao Scraper do?

The Actor searches Toutiao's public search surface and extracts the structured result data embedded in each response. It does not require a Toutiao login, a Chinese phone number, or an unofficial API key.

It supports:

• one to 20 keywords per run;
• article-only, video-only, or combined searches;
• a global result limit from 1 to 200;
• result deduplication by source ID or URL;
• normalized publication times and engagement counts;
• JSON, CSV, Excel, XML, RSS, and HTML exports through Apify datasets;
• scheduled runs for repeat monitoring.

Who is this Toutiao data for?

Media intelligence teams can watch how topics and organizations appear in Chinese news search.

Brand and communications teams can schedule recurring queries for company, executive, campaign, or product names.

Researchers and journalists can collect result headlines, snippets, publishers, and links for reproducible analysis.

Data engineers can send normalized rows to a warehouse, spreadsheet, webhook, or enrichment pipeline.

AI teams can use public result metadata as discovery input for translation, classification, clustering, and retrieval workflows.

Why use this Actor?

• Keyword workflow: provide search terms rather than collecting a known profile or URL.
• Articles and videos: exercise Toutiao's separate information and video search tabs.
• Transparent provenance: every row keeps the query, destination URL, host, and collection time.
• No false completeness: unavailable source fields are returned as null; they are never invented.
• Cost control: the output limit caps the number of saved records.
• HTTP-first design: no browser or automatic residential proxy fallback is used.

What data can you extract?

Field	Meaning
query	Keyword that produced the result
contentType	article or video search tab
rank	Position among parsed results for that query and tab
contentId	Stable result/content ID when Toutiao exposes one
title	Article headline or video title
url	Destination URL returned by Toutiao search
host	Destination hostname
source	Publisher, author, or source site when exposed
summary	Search-result abstract or snippet
publishedAt	ISO 8601 publication time when exposed
imageUrl	Primary image or video thumbnail when exposed
durationSeconds	Video duration when exposed
commentCount	Comment count when exposed
readCount	Article reads or video plays when exposed
likeCount	Like count when exposed
hasVideo	Whether the result is or contains video
scrapedAt	UTC collection timestamp

All source-dependent fields are nullable because Toutiao does not expose every field for every result.

How to search Toutiao in 3 steps

1. Open the Actor input page in Apify Console.
2. Add one or more keywords, select a content type, and choose maxItems.
3. Start the run and open the Dataset tab to inspect or export results.

A small article search:

{
  "keywords": ["人工智能"],
  "contentType": "articles",
  "maxItems": 10
}

A combined monitoring run:

{
  "keywords": ["人工智能", "新能源汽车"],
  "contentType": "both",
  "maxItems": 30
}

Input parameters

Parameter	Type	Default	Rules	Description
keywords	string array	['人工智能']	1–20 non-empty values	Terms searched independently
contentType	string	both	articles, videos, or both	Toutiao search tabs to collect
maxItems	integer	20	1–200	Global unique-result and charge cap

Duplicate keywords are removed. An empty keyword list, unsupported content type, or out-of-range limit fails the run with a clear error.

Example output

This shortened record reflects the current article-search output shape:

{
  "query": "人工智能",
  "contentType": "article",
  "rank": 1,
  "contentId": "7672133299032097320",
  "title": "坚定不移探索通用人工智能发展的中国道路",
  "url": "https://www.toutiao.com/group/7672133299032097320/",
  "host": "www.toutiao.com",
  "source": "光明网",
  "summary": "2023年4月28日，习近平总书记主持中共中央政治局会议…",
  "publishedAt": "2026-08-09T20:35:50.000Z",
  "imageUrl": null,
  "durationSeconds": null,
  "commentCount": 0,
  "readCount": 52,
  "likeCount": 2,
  "hasVideo": false,
  "scrapedAt": "2026-08-12T06:58:32.540Z"
}

Counts and ranking are snapshots of what the public search response exposed at collection time.

How much does it cost to search Toutiao articles and videos?

This is a pay-per-event Actor. Each run has one start event, and each successfully saved dataset row has one item event. Failed, malformed, duplicate, or rejected rows are not charged as items.

Current BRONZE pricing:

• start fee: $0.005 per run;
• result fee: $0.0124 per saved item.

At BRONZE rates, a 10-result run costs about $0.129, a 50-result run about $0.625, and a 100-result run about $1.245. Apify plan tiers may apply lower per-item rates. Compute or platform usage can appear separately according to your Apify plan.

The Actor stops as soon as maxItems records have been saved, so the input is also a simple spending control.

Monitoring and integration workflows

Scheduled media monitoring

Create an Apify Schedule with a stable keyword list. Store each run's dataset ID and compare contentId, url, or title across runs to identify newly surfaced results.

Spreadsheet export

Open a completed dataset, select CSV or Excel, and export the overview fields. Use query and contentType as grouping columns.

Webhooks and data warehouses

Attach a run-succeeded webhook, fetch dataset items from the API, then load them into BigQuery, Snowflake, PostgreSQL, Supabase, or another system.

Translation and classification

Send title and summary to your approved translation or NLP service. Keep url, source, and scrapedAt alongside derived labels for traceability.

Run the Actor with the Apify API

Replace YOUR_TOKEN with an Apify API token.

cURL

curl -X POST \
  "https://api.apify.com/v2/acts/automation-lab~toutiao-articles-videos-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"keywords":["人工智能"],"contentType":"articles","maxItems":10}'

JavaScript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('automation-lab/toutiao-articles-videos-scraper').call({
  keywords: ['人工智能'],
  contentType: 'both',
  maxItems: 20,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Python

from apify_client import ApifyClient

client = ApifyClient('YOUR_TOKEN')
run = client.actor('automation-lab/toutiao-articles-videos-scraper').call(run_input={
    'keywords': ['新能源汽车'],
    'contentType': 'videos',
    'maxItems': 10,
})
items = client.dataset(run['defaultDatasetId']).list_items().items
print(items)

Use Toutiao Scraper through MCP

Add the Apify MCP server in Claude Code:

claude mcp add --transport http apify \
  "https://mcp.apify.com?tools=automation-lab/toutiao-articles-videos-scraper"

Claude Desktop, Cursor, and VS Code

Claude Desktop, Cursor, and VS Code clients can use this MCP server configuration:

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com?tools=automation-lab/toutiao-articles-videos-scraper"
    }
  }
}

Example prompts:

• “Search Toutiao articles for 人工智能 and return the first 10 results.”
• “Collect article and video results for 新能源汽车 and summarize the publishers.”
• “Run my two Chinese brand keywords and save no more than 30 records.”

Tips for useful results

• Prefer specific Chinese terms over broad English terms.
• Put close variants in separate keyword entries rather than combining them into one long query.
• Start with 10–20 items while refining a monitoring query.
• Use articles when publisher and publication metadata matter most.
• Use videos to inspect Toutiao's federated video-search result set.
• Preserve the returned host: a video result may point to a third-party destination surfaced by Toutiao.

Limits and failure behavior

Toutiao controls search ranking, available result count, and exposed metadata. One current server-rendered result page is collected for each keyword and selected tab; requesting a larger maxItems does not manufacture more results.

The video tab is federated. It can return destinations on Bilibili or other sites in addition to Toutiao properties. host makes this explicit.

The Actor retries transient network, HTTP 429, and server failures with bounded backoff. If Toutiao returns a challenge or an unrecognized page instead of structured search cards, the run fails rather than reporting a misleading empty success.

No automatic proxy fallback is enabled. This keeps costs predictable, but access can vary if Toutiao changes geography or delivery rules.

Responsible use and legality

This Actor extracts publicly returned search-result metadata. You are responsible for ensuring that your use complies with Toutiao's terms, applicable copyright rules, privacy laws, database rights, and your organization's policies.

Do not use the Actor to collect sensitive personal data, bypass access controls, republish protected article bodies, or make automated decisions about individuals. Minimize retained data, honor deletion obligations, and preserve source attribution where required.

Troubleshooting

Why did I receive fewer rows than maxItems?

maxItems is a ceiling, not a promise. Toutiao may expose fewer valid cards for a keyword, and duplicates are removed. Try another specific Chinese keyword or add related keyword variants.

Why do video URLs point outside toutiao.com?

Toutiao's video search is federated and can surface third-party video destinations. Use host to filter destinations downstream.

Why did the run fail with “no search result data”?

Toutiao returned an unexpected or challenged page. Check Apify's status page, wait before retrying, and review the run log. Repeating the same request rapidly is unlikely to help.

Can I download article bodies or video files?

No. This Actor returns search-result metadata and destination links. It does not download media or extract full article bodies.

Can I search more than 20 keywords?

Split larger lists into multiple runs or Tasks. The per-run cap keeps requests bounded and makes schedules easier to audit.

FAQ

Does this require a Toutiao account?

No. It uses a public server-rendered search surface.

Does it support English keywords?

Yes, but Toutiao is a Chinese-language product and Chinese terms generally produce more relevant results.

Are article and video records charged differently?

No. Both are saved under the same item event and have no separate content-type charge.

Does the Actor use residential proxies?

No. The implementation is direct HTTP and has no hidden proxy fallback.

Are results deterministic?

The extraction is deterministic for a given response, but Toutiao ranking and inventory change over time.

Related Automation Lab Actors

• Google News Scraper for broader international news discovery.
• WeChat Official Account Articles Scraper for another Chinese publishing workflow.
• Youku Video Search Scraper for video-specific discovery on Youku.

Choose this Actor when the required source is Toutiao's keyword search and the desired output is public result metadata.