WeChat Official Account Articles Scraper

📰 What is WeChat Official Account Articles Scraper?

WeChat Official Account Articles Scraper extracts articles of WeChat Official Accounts (公众号): title, summary, account, publication date and cover for every article, and with Full article on the full text, Markdown, images, author, permanent link, original badge (原创), IP region, hashtags and album. Search by keyword, paste article or album links, or list accounts. It works as a WeChat API alternative: run it on a schedule, call it from your code, or plug it into Make, Zapier or n8n — for media monitoring, market research in China, brand tracking or AI datasets.

Type a keyword (人工智能) or paste WeChat links, click Start, and download the articles in JSON, CSV or Excel. No WeChat account needed, nothing to set up, and it is fast: about 90 articles from a keyword in about 40 seconds, or 86 full articles (text, images, permanent link) in about 5 minutes.

📋 What data can you extract from WeChat?

One item per article (plus optional account rows and the rows of the data options), 63 fields in all, 56 of them on an article:

Category	What you get
📰 Article	title, summary, cover image, publication date and time, article ID — 锂电池及新能源车板块，齐跌！
🔗 Links	permanent link that never expires, temporary search link, "Read the original" link
📝 Content	full text (one paragraph per line), Markdown, raw HTML, text length, content type (article, video, audio…)
🖼️ Media	every image of the body with its width and height, image count, embedded video IDs
✍️ Author and badges	author's byline, original badge (原创), paid article, IP region and country of publication — 广东
🏢 Account	account name, WeChat ID, original gh_ ID, biz, verified, subscription or service account — 中国基金报
👤 Account profile	description, avatar, number of its articles in the run, its newest article (one row per account)
📚 Album (合集)	album title, link, size and the article's position in it — 港股日报, 463 articles
🏷️ Topics	WeChat hashtags of the article; your own free labels (matchedLabels) from the keywords you give
🔎 Search context	keyword, results page and rank, estimated number of results, where the article was found
📈 Data options (paid, off by default)	the WeChat Index of your keywords over 30 days, and the search suggestions WeChat gives for them

Every field, with an example, is listed in the Output section below.

The search-card fields (title, summary, account name, date, cover, link) are always there — except the summary of an album line, which comes with Full article only. The text, images, author, permanent link, badges, IP region, hashtags and album are filled when Full article (extractDetails) is on — off by default, charged apart. Read counts, likes and comments are not shown by WeChat outside its app: they are never returned as fake zeros.

✅ Why use WeChat Official Account Articles Scraper?

• 🚀 Fast: about 90 articles from a keyword in about 40 seconds; 86 full articles with text and images in about 5 minutes.
• 🧩 Five ways in, one run: keywords, article links, album links, WeChat search links and account names, all at once — no mode to choose.
• 🔗 Permanent links: articles of an album, and pasted links of any kind (short, temporary or search links), come back with the link that never expires and their IDs.
• 📚 Whole albums (合集): paste an album link, or any article of it, and get every article of the album, newest first.
• 📝 Ready for AI and analysis: full text, Markdown with headings and images kept, every image with its size.
• 🔔 Monitoring built in: tick Only new articles, schedule the Actor, and each run returns (and charges as articles) only what it has never delivered before.
• 📅 Date filters: postedAfter: "7 days" — filtered-out articles are not saved, never count in maxItems, and their page is not opened.
• 🔌 API, scheduling, monitoring, integrations (Make, Zapier, n8n, Google Sheets…) and JSON/CSV/Excel export via the Apify platform.

🚀 How to scrape WeChat articles

1. Create a free Apify account.
2. Open WeChat Official Account Articles Scraper and type a Search keyword (e.g. 人工智能), Chinese or English.
3. Or fill any of the other inputs: Article links, Album (合集) links, Sogou WeChat search links or Official accounts — they all add up in one run.
4. Set Max articles (20 in the form, 100 by default through the API, 0 = no limit), tick Full article if you want the text and images, then click Start.
5. Download the dataset in JSON, CSV, Excel or via API.

💰 How much does it cost to scrape WeChat?

This Actor uses pay per event pricing: you pay for the articles you get, and only for the options you tick. The prices of each event, for every Apify plan (higher plans pay less), are in the Pricing tab:

• Article (result): every article row saved — from a search, an album, an account, a pasted link or the history.
• Full article (article-content), on top of the article: an article page read with Full article on (text, images, permanent link…). A row without content (deleted page, check page) is never charged as a full article.
• Link conversion (link-conversion), on top of the article: a pasted article link with Full article off, turned into its permanent link and IDs, without the text.
• Account profile (account-profile): one account row, with Account profiles on. Account rows never take the place of an article in Max articles.
• Search statistics (market-stats): one sheet of figures per search, with Statistics per search on, charged once per sheet at the end of the run. A search with no article delivered has no sheet and costs nothing.
• WeChat Index (wechat-index): one row per keyword, with WeChat Index on; charged for each keyword looked up, also when WeChat has no data for it (no row then).
• Search suggestions (search-suggestions): one row per keyword, with Search suggestions on; charged for each keyword looked up, also when WeChat has no suggestion for it (no row then).
• History article (history-article), on top of the article: an article read from an account's full history, with the optional WeChat login.
• Filter check (filter-check): an article read and dropped by one of the Actor's own filters (date or keywords), or, with Only new articles, an article already delivered that the run reads again (once per article and per run). Lower than an article — filtering never costs more than taking everything.
• A small fee per run start (apify-actor-start).

Platform usage (compute, proxy) is included in the price. Every option is off by default: tick only what you need. The two data options (WeChat Index, Search suggestions) are for a paid Apify plan only: on the free plan they are skipped, with a message in the log, and nothing is charged.

⚙️ Input

{
    "query": "人工智能",
    "maxItems": 100
}

Several keywords, the full text, a cap per keyword, only recent articles, only the ones not delivered before:

{
    "searchQueries": ["人工智能", "区块链"],
    "maxItemsPerQuery": 50,
    "extractDetails": true,
    "postedAfter": "7 days",
    "onlyNew": true,
    "stateKey": "ai-news"
}

Or with links and accounts, all in one run:

{
    "articleUrls": ["https://mp.weixin.qq.com/s/4gG0wmpjHS9"],
    "albumUrls": [
        "https://mp.weixin.qq.com/mp/appmsgalbum?__biz=MzA5MDEzNjQwMA==&action=getalbum&album_id=3080785052776808449"
    ],
    "searchUrls": ["https://weixin.sogou.com/weixin?type=2&query=%E4%BA%BA%E5%B7%A5%E6%99%BA%E8%83%BD&ie=utf8"],
    "accounts": ["中国基金报"],
    "extractDetails": true,
    "includeAccountProfiles": true,
    "maxItems": 500
}

Field	Notes
query, searchQueries	Article search, Chinese or English, exactly as typed in WeChat (人工智能); searchQueries adds more keywords, one search each. WeChat shows about 100 articles per keyword: add keywords to get more.
articleUrls	Article links of any form: permanent, short (https://mp.weixin.qq.com/s/4gG0wmpjHS9), temporary or redirect links from WeChat search. With extractDetails the whole article; without, a link conversion (permanent link when WeChat reveals it, IDs, account, title, date, no text). Up to 1,000 links.
albumUrls	Album (合集) links (https://mp.weixin.qq.com/mp/appmsgalbum?__biz=MzA5MDEzNjQwMA==&action=getalbum&album_id=3080785052776808449) or an article link of the album: every article of the album, newest first, with its permanent link. Up to 200 albums.
searchUrls	Search-result links copied from weixin.sogou.com (https://weixin.sogou.com/weixin?type=2&query=%E4%BA%BA%E5%B7%A5%E6%99%BA%E8%83%BD&ie=utf8); pagination is automatic.
accounts	Account names (中国基金报), WeChat IDs (chinafundnews) or an article link of the account. Without login: the account's most visible articles (about 100 at most). With the login below: its full history.
maxItems, maxItemsPerQuery	Stop after this many articles for the whole run, and for EACH keyword, search link, album or account (0 = no limit). Keywords, search links, albums and accounts add up (max 500 searches per run).
maxPagesPerQuery	Search-result pages read per keyword or search link, 10 articles each (10 = all that WeChat shows).
extractDetails, includeMarkdown, includeHtml	Full article (off by default, charged apart): text, author, images, permanent link, badges, IP region, hashtags, album. Markdown (on, free) and raw HTML (off, free, large) of the body come with it.
includeAccountProfiles	One more row per account that has an article in the results: description, avatar, IDs, verified, number of its articles in the run, newest article. Charged per account.
includeMarketStats	One sheet of figures per search (keyword, album, account or pasted Sogou link), in the MARKET_STATS record of the key-value store: articles delivered and results announced, first and last date, articles per day (Beijing time, UTC+8), accounts and the top 10 accounts, shares of original and paid articles, article types, articles with images, average text length, top 10 hashtags, IP regions. Counted on the articles of this run only; each share comes with the number of articles that give the value (most page fields need extractDetails). Off by default, charged per sheet.
includeWechatIndex	WeChat Index (paid option, paid Apify plan only): one row of type wechat-index per keyword of query / searchQueries with its WeChat Index, one value a day over the last 30 days (Beijing calendar), the latest value and the change over the period. Charged per keyword looked up, also when WeChat has no data for it.
includeSearchSuggestions	Search suggestions (paid option, paid Apify plan only): one row of type search-suggestions per keyword with the words WeChat search suggests, in its order. Charged per keyword looked up, also when WeChat has no suggestion for it.
suggestionScope	With Search suggestions: the part of WeChat search the suggestions come from (ALL by default, or ACCOUNT, ARTICLE, CHANNEL, MOMENTS, LIVE, EMOJI, AUDIO, BOOK, WECHAT_INDEX, NEWS, MINI_PROGRAM, ENCYCLOPEDIA).
postedAfter, postedBefore	Publication date range: 2026-09-01, or a period before now (7 days, 2 weeks, 1 month; via the API also 24 hours).
excludeKeywords	Drop the articles whose title or summary (the search snippet, or the digest the account wrote) contains one of these words (case ignored).
keywordLabels	Free (no extra charge, nothing dropped): your own labels on the articles. One line per label, Tech: AI, chip, 芯片 adds Tech when at least one of the words is found in the title, the summary or the text; a line with no colon (blockchain) is one word that is its own label. Words are separated by a comma, an English one or a Chinese one (， or 、). Case, accents and full-width letters are ignored, and a word is found anywhere in the text (so it works for Chinese). The text needs Full article: without it only the title and the summary are searched. Up to 50 labels of 50 words. Result: matchedLabels.
onlyNew, stateKey, resetState	Monitoring: only the articles never delivered under this memory key (ai-news); resetState forgets the memory.
mpCookie, mpToken	Optional login with your own Official Account (a free subscription account is enough), only for the full history of the accounts in accounts. Stored encrypted, never written to the log or the results.
Advanced	proxyConfiguration (Apify proxy by default, included in the price; the residential proxy is not available), maxConcurrency, maxRequestsPerMinute, minRequestIntervalMs, maxRequestRetries, debugLog.

📦 Output

One full article (extractDetails on), found through its album, from a real article, shortened (long texts cut, some fields left out):

{
    "type": "article",
    "id": "3090136400_2656415723_3",
    "resultKey": "rk_3f1a9c0e5b7d2a14",
    "url": "https://mp.weixin.qq.com/s?__biz=MzA5MDEzNjQwMA==&mid=2656415723&idx=3&sn=35ee2516bca63312c69fb7a061e88f6c&chksm=8ab5c4eade9f43e03e91b4f5c9a25ab61d05ad42229fa6fa61129a728d3a87d306343d8107a6",
    "permanentUrl": "https://mp.weixin.qq.com/s?__biz=MzA5MDEzNjQwMA==&mid=2656415723&idx=3&sn=35ee2516bca63312c69fb7a061e88f6c&chksm=8ab5c4eade9f43e03e91b4f5c9a25ab61d05ad42229fa6fa61129a728d3a87d306343d8107a6",
    "title": "利好来了，五一视界大涨！",
    "summary": "中国基金报微信：Chinafundnews",
    "author": "伊万",
    "publishedAt": "2026-10-20T09:49:41.000Z",
    "publishedTimestamp": 1792489781,
    "accountName": "中国基金报",
    "accountId": "gh_a4a87df43a4c",
    "accountWechatId": "chinafundnews",
    "accountBiz": "MzA5MDEzNjQwMA==",
    "accountType": "subscription",
    "contentType": "article",
    "content": "【导读】港股收跌，创新药板块走强，翰森制药大涨超8%，物理AI及机器人概念股五一视界大涨超22%\n中国基金报记者 伊万\n…",
    "contentLength": 1166,
    "images": [
        {
            "url": "https://mmbiz.qpic.cn/mmbiz_png/7KcgpBE2JbqDNAMxtibciaSG3HZoxLydv6bsA8NibTTB6Ac7rBBJFMVaGq3wy9KCjNcBBsib2tbyYUUWdbibTeXXLrw/640?wx_fmt=png&from=appmsg",
            "width": 1920,
            "height": 755
        }
    ],
    "imageCount": 7,
    "videoIds": [],
    "isOriginal": false,
    "isPaid": false,
    "ipRegion": "广东",
    "ipCountry": "中国",
    "hashtags": [],
    "matchedLabels": [],
    "albumId": "3080785052776808449",
    "albumTitle": "港股日报",
    "albumUrl": "https://mp.weixin.qq.com/mp/appmsgalbum?__biz=MzA5MDEzNjQwMA==&action=getalbum&album_id=3080785052776808449",
    "albumArticleCount": 463,
    "messageId": "2656415723",
    "indexInMessage": 3,
    "source": "album",
    "searchQuery": "港股日报",
    "contentStatus": "ok",
    "scrapedAt": "2026-10-21T12:00:00.000Z"
}

You can download the dataset in various formats such as JSON, HTML, CSV or Excel.

All 56 fields

Fields	What you get
type, id, resultKey, docId	article or account (the rows of the data options are wechat-index or search-suggestions, below); 3090136400_2656415723_3 (the ID WeChat albums use, when known); a stable key, the same whatever the source and across runs; the ID in WeChat search
url, permanentUrl, temporaryUrl, sogouUrl	Links: the best link available; the permanent link (sn + chksm) when WeChat reveals it; the signed search link, valid a few hours; the search redirect link
title, summary, coverUrl	Article: 利好来了，五一视界大涨！, search snippet or digest, cover image. Without Full article, a title found by search is written as Sogou prints it (half-width signs: , : ! for ， ： ！); with it, as the account wrote it
publishedAt, publishedTimestamp	2026-10-20T09:49:41.000Z (what the date filters read), Unix seconds
author, isOriginal, isPaid, sourceUrl	Author and badges: 伊万, original badge, pay-to-read article (only the free part is returned), "Read the original" link
ipRegion, ipCountry	广东, 中国: where WeChat says the article was published from
accountName, accountId, accountWechatId, accountBiz, accountType	Account: 中国基金报, gh_a4a87df43a4c, chinafundnews, MzA5MDEzNjQwMA==, subscription or service
accountDescription, accountAvatar, accountVerified	Account profile (account rows): 实时专业理财资讯, avatar URL, true
articlesFound, latestArticleTitle, latestArticleAt, latestArticleUrl	Account rows: number of its articles found in this run, its newest article (title, date, link)
contentType, content, contentMarkdown, contentHtml, contentLength	Content: article, video, audio, image or text; full text; Markdown; raw HTML; 1166 characters
images, imageCount, videoIds	Media: every image with url, width, height; 7; IDs of the embedded videos
hashtags	WeChat topic tags
matchedLabels	Keyword labels (free): the labels of your keywordLabels found in the title, summary or text (["Tech"]), in the order you wrote them, once each; [] when none, and on account rows
albumId, albumTitle, albumUrl, albumArticleCount, albumPosition	Album: 3080785052776808449, 港股日报, album link, 463 articles, 1 = newest
messageId, indexInMessage	2656415723, 3 (position of the article in its push, 1 = headline)
source, searchQuery, searchPage, searchPosition, totalResultsEstimate, searchUrl	Search context: search, article, album or account; the keyword, album or account that found it; results page and rank; results WeChat announces; the results page it was found on
contentStatus, scrapedAt	ok, not_requested (Full article off), deleted, verification (WeChat asked for a check twice), paid or error; ISO timestamp

Rows of the data options (7 more fields)

A row of a data option has the fields type, id, resultKey, scrapedAt and its own fields below, nothing else. These rows are always delivered, also with Only new articles: the figures change every day.

Fields	What you get
keyword	All data rows: the keyword of query / searchQueries the row answers
title	WeChat Index: the title of the curve (人工智能 - 微信指数)
latestIndex, latestDate, dataPoints, changePercent	WeChat Index: the index of the last day (69931366), that day (2026-10-23, Beijing calendar), one {date, index} per day over about 30 days, and the change from the first to the last day in percent
scope, suggestions	Search suggestions: ALL or the part of WeChat search asked, and the suggested words (["人工智能", "人工智能训练"])

💡 Tips

How to get more results

WeChat shows about 100 articles per keyword (10 pages of 10). To get more, add keywords to searchQueries — synonyms, names, Chinese and English — or paste album links: an album returns all its articles, not only the first 100. Set maxItems to 0 for no global cap.

How to reduce costs

The price is per article, so the levers are maxItems, maxItemsPerQuery, the filters (a filtered-out article costs only the filter fee, lower than an article, and its page is not opened) and onlyNew for recurring runs (an article is never saved twice; each one a run reads again costs one filter check, as Sogou's results are ordered by relevance, not by date). Leave Full article off when titles, summaries, dates and links are enough.

Several searches in one run

Fill searchQueries, searchUrls, albumUrls and / or accounts: each keyword, search link, album and account is a search of its own, up to 500 per run. The single query field still works and is added to the list. An article found by several searches is saved — and charged — once. Set maxItemsPerQuery to give every search its own cap: without it the first searches can use up the whole maxItems budget. Pasted articleUrls are not searches: each link gives its article.

Monitoring: only the new articles

Tick Only new articles (onlyNew) and schedule the Actor. The first run returns everything; each later run skips the articles already delivered: they are not saved again, each one a run reads again costs one filter check, and for articles found by search, album, account or history their page is not even opened. The memory lives in a named key-value store of your account (wechat-articles-scraper-seen, up to 150,000 articles per key) and is only updated with articles that really reached the dataset, so a failed run never hides anything. Give each schedule its own stateKey (two schedules sharing a key would hide each other's articles), and tick resetState once to start over. Albums and account histories are read newest first and stop once they meet a long run of articles you already have; keyword searches are ranked by relevance, so they are read to the end.

Filter by publication date

postedAfter and postedBefore take a date (2026-09-01, the whole day is included, Beijing time) or a period before now (7 days, 2 weeks, 1 month; via the API also 24 hours or a full ISO date-time). The filter reads publishedAt on the search card, album line or history entry, before the article page is opened. Filtered-out articles are not saved and do not count in maxItems (each article dropped costs the filter fee, see Pricing); the run summary tells how many were filtered.

WeChat Index and suggestions

Tick WeChat Index and / or Search suggestions to get, for each keyword of query and searchQueries, its popularity over 30 days and the words WeChat suggests. They are paid options, on a paid Apify plan only, and use no article: with maxItems at a low value you can run them for the data rows alone. If the data service is out of reach for a moment, the run is not failed: a line in the log says the option was not delivered. A lookup is charged when it is started, also when it gets no data. The run's maximum cost is respected: an option stops when it can no longer be paid.

Full history of an account

Without login, accounts returns the articles of the account that WeChat search shows (about 100 at most). For its full archive, log in to mp.weixin.qq.com with your own Official Account (a free subscription account is enough), copy the cookie and the token into mpCookie and mpToken, and list the accounts in accounts (gh_ IDs are accepted too). Each history article is charged as a history article on top of the article. A login WeChat refuses or that has expired is reported in the log with a clear message: log in again and copy a fresh cookie and token.

🔌 Integrations and API

Call the Actor via the Apify API, the JavaScript or Python clients, or connect it with integrations and webhooks (Make, Zapier, n8n, Google Sheets, Slack, Airtable…). The dataset can be fetched as JSON or CSV from any tool.

🤖 Use with AI agents (MCP)

AI agents (Claude, ChatGPT, Cursor…) can find and run this Actor through the Apify MCP server, billed to their Apify account like any run. It returns one item per WeChat Official Account article. Actor id: nice_dev/wechat-articles-scraper; MCP server with this Actor only: https://mcp.apify.com/?tools=fetch-actor-details,nice_dev/wechat-articles-scraper.

Smallest input, for a cheap first call:

{
    "query": "人工智能",
    "maxItems": 10
}

Key output fields: url, title, summary, accountName, publishedAt, and with extractDetails content, contentMarkdown and permanentUrl.

Cost: pay per event — per article, plus the options ticked (full article, link conversion, account profile, WeChat Index, search suggestions), see the pricing section above and the Pricing tab. Cap each call with maxItems and, through the API, with the run option maxTotalChargeUsd.

❓ FAQ

Is it legal to scrape WeChat?

The Actor only reads articles that WeChat Official Accounts publish openly, which anyone can open without a WeChat account. It logs in to nothing, except the optional history mode, which uses your own login. Results can contain personal data (author bylines, account avatars, IP region of publication), which is protected by laws such as GDPR in Europe and PIPL in China: do not store it without a legitimate reason. You are responsible for using the data in compliance with WeChat's Terms of Service and applicable law; if in doubt, ask a lawyer. This Actor is not affiliated with WeChat or Tencent.

Do I need a WeChat account?

No. Search, articles, albums and accounts work without any account. The login (mpCookie + mpToken) is optional and only adds the full history of the accounts you list.

How many articles can I get per keyword?

About 100: WeChat search shows 10 pages of 10 articles per keyword, and some come back twice, so a keyword usually gives 85 to 95 articles. Add keywords, albums or accounts to collect more.

Does it need a proxy?

The proxy is included in the price: leave the default setting (the residential proxy is not available). A request the site turns away is retried at once on a new proxy session, up to 10 times on top of the retries (without a proxy, after a pause of 5 seconds, doubled at each retry up to 150 seconds).

Is the data safe to open in Excel or to show on a web page?

Titles and texts are the accounts' own words, copied as they are. A text can begin with -, +, = or @: Excel and Google Sheets may read such a cell of a CSV file as a formula or as a number. The Actor leaves the text as it is, so that the JSON and the API give the real value: when you open a CSV, import these columns as text. contentHtml is the article's own HTML, not sanitized: on a web page, escape every field like any text written by a stranger.

Known limitations

• About 100 articles per keyword, and an account without login gives its most visible articles, not its full archive.
• Read counts, likes and comments are only shown inside the WeChat app: they are not returned.
• A pasted article link without extractDetails is a link conversion: permanent link when WeChat reveals it, IDs, account, title and date, but no text or images.
• A pasted article link is always opened, even with onlyNew or the filters on: it has to be read to be recognised. An article already delivered or filtered out is then not saved and not charged as an article (it costs one filter check).
• permanentUrl is null for a search result that belongs to no album and was not opened through a permanent link: WeChat does not reveal it. url then holds the temporary link, valid a few hours.
• A few article pages answer with a WeChat check page: the row is saved with contentStatus: verification and is not charged as a full article.
• WeChat Index and search suggestions: paid Apify plan only; the keywords of query and searchQueries only (not the accounts or links); a keyword WeChat has no index for gives no row (the lookup is still charged).
• Account profiles (includeAccountProfiles) are read for the accounts of the articles delivered. An account whose profile could not be read in time gets no account row and is not charged.
• onlyNew remembers article keys, not their content: an article edited since is not returned again.
• Two articles of the same account published the same day (Beijing time) under the same title share one resultKey: only the first is saved and charged.
• Two runs sharing the same stateKey at the same time may both return the same new article.

A run that reaches its timeout stops itself about 45 seconds before it: no new page is asked, what it read is saved and, with onlyNew, remembered, and the run ends Succeeded with "Stopped before the run's timeout". Resurrect it to go on from there, or give the next run a longer timeout (Run options). The account rows are written at that stop and not updated by the resurrected run: their articlesFound and latest article are those of the first part.

A run the platform stops without warning (out of memory)

• Resurrect it: it goes on from where it stood at most a minute before the stop. What it had read since is read again, and the articles already saved are skipped: none is delivered or charged twice, and maxItems still counts them.
• With onlyNew, the memory is saved after each write (at most once every 2 seconds): resurrect the stopped run and the articles it had saved meanwhile join the memory; leave it stopped for good, and the next run may return the articles of its last 2 seconds once more.
• After such a stop (or a migration of the run to another server), the articlesFound count of the account rows may be lower than the number of articles of the account in the dataset.

Something doesn't work?

The last line of the log counts the articles saved and filtered out, and the requests that failed after every retry. Those requests are listed, with the reason, in the FAILED_REQUESTS record of the run's key-value store. A run that saved nothing and had failed requests fails, and its last message gives the cause.

If WeChat changes its pages, you are told instead of paying for blank rows. A results page that counts articles but gives none the Actor can read is an error (listed in FAILED_REQUESTS), never a quiet "No articles found". If the first 20 articles read all lack their title, publication date, account name or — with extractDetails — text, the run saves nothing more, stops and fails, and its last message names the missing field: at most those first articles are charged. An article that postedAfter / postedBefore drops because it has no date at all counts among those 20. A run of fewer articles (a scheduled onlyNew run that finds a few new ones, a small Max articles) is checked at its end, from 5 articles: its articles are already saved, and the run fails with the missing field instead of ending green.

🛟 Support

Open an issue in the Issues tab with a link to your run: the run log and the FAILED_REQUESTS record of the key-value store show exactly which URLs failed and why.