AI agents: read https://whizbrew.com/llms.txt (plain text: what Whizbrew is, what a call costs, how to start). This page as Markdown: https://whizbrew.com/api/x-twitter-data-api.md, or send Accept: text/markdown. Prices as data: /manifest.json.

WhizbrewToolsAPIsPricingDocsStatusStart free

X (Twitter) data API for AI agents: from $0.002 per call ($2 per 1,000), no subscription

Whizbrew's X (Twitter) data API costs from $0.002 per call ($2 per 1,000 calls); a typical call is $0.003 ($3 per 1,000). Only a call that returns data is charged: an empty or failed call is free. There is no subscription and no monthly fee. A new account starts with $2 of free credit, with no card.

It reads X profiles, posts, post search and post history: 18 tools, live at the moment of the call, over one API key or one MCP server (https://whizbrew.com/mcp). It is an API, not a scraper you run: no proxies, no browser and no scraping code on your side. The same key reaches 893 tools for other platforms (all APIs, all tools, pricing).

Pricing

ToolWhat it returnsPer call that returns dataPer 1,000 calls
x.profilePublic X profile and its counts, by handle or by numeric user id. Give one of handle or id.$0.003$3
x.postsPosts of an X account, newest first, with exact views, likes, replies, reposts, quotes and bookmarks, the language, the author's name and account id, the thread (conversation) id on the rows of X's pick, and the media of each post. X shows a visitor without a login the pinned post and about 4 of the account's newest posts (feed: newest); the rest of the list is X's own pick of the account's posts, often older (feed: x_pick). Set newest_only to leave the picks out.$0.003$3
x.postOne public X post by its link: text, author, counts and views.$0.002$2
x.repliesThe top replies to a public X post. Only the few X shows without a login, not the full list.$0.003$3
x.communityPublic X community: name, description, member count, join policy and rules.$0.002$2
x.community_postsPosts of a public X community in X's ranked order (one page, up to 20).$0.003$3
x.trendsTrending topics on X right now, in X's rank order, worldwide or for a country or city.$0.002$2
x.articleA long-form X article from its post link: title, full text and text blocks.$0.003$3
x.threadThe author's own thread around an X post: the first post and the author's follow-up posts, in order.$0.004$4
x.transcriptThe captions X made for the video of a post, as text with times. Only when X has captions for it.$0.004$4
x.newsThe news stories X shows right now: title, category and how many posts each has.$0.002$2
x.storyOne X news story: summary, text, its top posts and related stories.$0.003$3
x.historyThe posts of an X account, page after page, newest first: each with its date, text, kind (original, reply, quote, repost), media type and the counts (views, likes, replies, reposts, quotes, bookmarks), and "coverage" saying how complete the list is and why. Free route (default): an account X lists as an organisation (a company, a government body) gives about its newest 800 posts and reposts in date order, about 25 a page, paged with next_cursor, without its replies to other accounts; a personal account gives only X's own pick of about 100 posts and no paging (coverage.mode: x_pick). since and until keep the posts of a date range; a range far back costs the pages before it. source: exact reads every post back to the first through a paid route, 100 posts a page, and is not switched on yet.$0.003; with source = exact: $0.52$3
x.post_statsWhat works best and worst for an X account and how its numbers moved over time, computed from its posts (counts and medians, no model): posts and median views per week or month, the trend with its method, the 5 best and 5 worst posts, and the performance by kind (original, reply, quote, repost), by media type, by weekday, by hour, by text length, plus the words that are more common in the best quarter of posts than in the worst. Give a handle (the posts are read with x.history: see its limits, "input.coverage" repeats them) or your own list in data.posts: rows of x.history or x.posts, the tweets.js of the archive an account owner downloads from X, or posts fetched elsewhere. Says how many posts it rests on.$0.004$4
x.searchX posts for a keyword, with exact counts: X's own search, about 20 posts a page, each with its text, date, author (with followers and verified mark), exact likes, reposts, replies, quotes, views and bookmarks, language, media, and what it replies to, reposts or quotes. sort: relevance (X's Top tab) or newest (X's Latest tab, newest first). time_range keeps to posts of the last day, week, month or year; next_cursor reads the next page. X's search operators work in the query (from:nasa, "exact phrase", lang:en, min_faves:100). X matches more than the post's text, so a row can lack the keyword in text. The data comes from a licensed social index (source says so, with as_of). When the licensed index is not available the tool answers from public search-engine indexes instead (source: search_index): a few dozen posts at most in the engines' order, not fresh, no cursor, and found_by, index_* and engines_* describe that route. A keyword with no post is free.$0.004$4
x.timelineThe posts and reposts of an X account as its profile lists them to a signed-in visitor, newest first, about 20 a page and page after page with next_cursor: text, date, exact likes, reposts, replies, quotes, views and bookmarks, language, media, and what each reposts or quotes. account has the account's id, name, followers and post count. Goes past what a visitor without a login is shown (x.posts: about 5 newest posts plus X's pick; x.history: about 800 posts for organisations only, free). Replies to others are not in this list. From a licensed social index (source, as_of). An account without posts is free.$0.003$3
x.followersThe accounts that follow an X account, about 70 a page: handle, name, id, followers, bio, location, website, verified mark and the date the account was made. total is the account's follower count; next_cursor reads the next page. From a licensed social index (source, as_of). An account without followers is free.$0.003$3
x.followingThe accounts an X account follows, about 70 a page: handle, name, id, followers, bio, location, website, verified mark and the date the account was made; next_cursor reads the next page. From a licensed social index (source, as_of). An account that follows nobody is free.$0.003$3

Example calls

x.profile ($0.003 per call): Public X profile and its counts, by handle or by numeric user id

curl -s -X POST https://whizbrew.com/v1/call/x.profile \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"handle":"duolingo"}'

x.posts ($0.003 per call): Posts of an X account, newest first, with exact views, likes, replies, reposts, quotes and bookmarks, the lang

curl -s -X POST https://whizbrew.com/v1/call/x.posts \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"handle":"duolingo","limit":3}'

x.post ($0.002 per call): One public X post by its link: text, author, counts and views

curl -s -X POST https://whizbrew.com/v1/call/x.post \
  -H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
  -d '{"url":"https://x.com/elonmusk/status/2121763396190179414"}'

Over MCP the same call is call_tool with {"tool": "<name>", "arguments": {…}}. How often each tool works, measured from real calls: https://whizbrew.com/status.

Start