web-search-and-content

Extract clean title, host, and bounded main text from one public URL for $0.01 USDC. Call when you need readable page content for routing, RAG prep, or summarization—not general web search and not a multi-URL crawl. Returns structured page metadata via x402. No API key, no subscription.

Extract forms, methods, actions, and fields for browser automation and workflow planning for $0.01 USDC. Call when you need this bounded result without an API key or subscription. Returns machine-readable JSON via x402 on Base or Solana. Limitation: results cover only the supplied input or one bounded public endpoint at request time; callers must verify before acting.

Renders a webpage exactly as a real browser would, including all its scripts and dynamic content, then takes a picture of it. You can capture just the visible area or the whole scrollable page, choose the screen size, wait for slow-loading images to finish, and hide cookie banners or popups before the picture is taken.

PDF text extraction API for agents: extract bounded UTF-8 text from one caller-supplied PDF for $0.003 USDC via x402. Local first-party parsing with SHA-256 receipt and no document storage or URL fetch; scanned-image OCR is a separate product.

Harvest URL and DOI citation patterns from text without network access — call when harvesting URL/DOI citation patterns without network. Use on text/URLs the agent already holds—adjacent to high-volume extract demand, never advertised as general web or X search. Returns deterministic machine-readable JSON for $0.003 USDC via x402 on Base. Execution is first-party, local-only, stateless, and memory-only with no paid upstream, no input retention, and no claim of live chain tip, web search, or med…

Strip HTML tags from caller-held markup. Call when you need plain text from agent-held HTML snippets without a browser. Returns tag-stripped text and length for $0.001 USDC via x402 on Base. No network, no keys, no retention; first-party local JSON only. Results are advisory; the caller owns budgets, auth, and production controls.

Extract scheme://host[:port] origin from a URL without network access — first-party local utility for agent preflight and proven micro-utils expansion. Returns deterministic machine-readable JSON for $0.001 USDC via x402 on Base. Execution is first-party, local-only, stateless, and memory-only with no paid upstream, no input retention, no web search, no live RPC, and no mediagen. Results are advisory; the caller owns authorization, budgets, and production controls.

Fetch a URL and extract clean main-content text with title, description, word count, and char count; boilerplate (nav/footer/sidebar/ads) is stripped.

Extract structured elements: heading hierarchy (h1-h6), lists, tables as 2D arrays, and images with alt text, plus element counts.

Perform a Google search and return structured results. Returns title, URL, snippet, position for each result. Google search SERP results for AI agents: organic results as clean JSON, scoped by country (ISO 3166) and language (ISO 639-1), up to 20 results per call.

Extract a public HTTP(S) web page into structured JSON with a clean text excerpt for LLM workflows: title, description, JSON-LD, Open Graph/Twitter metadata, headings, links, and AI-readiness signals. Fetches without JavaScript rendering, follows redirects, and applies a 12-second timeout and 3 MB read cap. The paid JSON keeps requestedUrl, finalUrl, source HTTP status, sourceOk, a nullable error, and capture limits; the excerpt is not the full page. Use /read for longer cleaned Markdown.

Extract a public HTTP(S) web page into structured JSON with a clean text excerpt for LLM workflows: title, description, JSON-LD, Open Graph/Twitter metadata, headings, links, and AI-readiness signals. Fetches without JavaScript rendering, follows redirects, and applies a 12-second timeout and 3 MB read cap. The paid JSON keeps requestedUrl, finalUrl, source HTTP status, sourceOk, a nullable error, and capture limits; the excerpt is not the full page. Use /read for longer cleaned Markdown.

Fetch any RSS or Atom feed and get normalized JSON items — title, link, ISO-8601 date, plain-text summary, author, categories. Point it at a regular HTML page and it auto-discovers the feed. RSS 2.0, RSS 1.0/RDF and Atom, messy real-world feeds included. The monitoring primitive for agents that watch sources.

Fetch any public URL and return clean, structured content: title, meta description, main text, markdown, headings, outbound links and page metadata. Removes scripts, navigation and boilerplate. Ideal for reading articles, docs and product pages before reasoning over them.

Convert PDF to text for AI agents - pass any public PDF URL, get the full extracted plain text in one call. pdftotext / convertpdftotext / pdf to text converter / extract text from PDF / read a PDF / parse PDF documents, papers, SEC filings, reports, invoices, manuals. Up to 25MB and 500k chars, with page count and a likely_scanned flag (text-based PDFs only, no OCR). No API key, no account. ?url=<public PDF url>. Free live sample: /pdf/example.

What is the web saying right now? Live search results condensed into clean text for agent reasoning — answers beyond any model's training cutoff. Google-grade when the node's search provider is configured.

Fetch any public web page or API and get it back as clean JSON — status, content type, final URL after redirects and the body as text, size-capped. The mid-task primitive agents use to read a doc, open a link or pull...

Voice-of-customer / product complaints digest for AI agents - pass a product or brand, get clustered pain points, feature requests and praise mined from live X posts, with verbatim quotes and URLs. Product research, competitor research, churn signals, roadmap input. ?product=<name or handle>. Raw tweets + trend summary instead: /x/digest $0.05.

Web search with the page text already extracted: one call returns the top results for a query AND the readable content of each result page (JavaScript rendered, markdown or text, truncated to max_chars). What an agent needs to answer a question with sources without a second round of fetches. Up to 5 pages per call, fetched in parallel.

Top AI News — HIGH-SIGNAL only: model/weights releases, security incidents, agents & harnesses, the major labs, legislation, launches/deals; marketing, how-tos, earnings, and PR filtered out. Newest first, `since` to poll, `limit` to page; no search.

Page speed API - Google PageSpeed Insights (Lighthouse) performance audit for a URL on mobile or desktop. Returns the performance score plus lab Core Web Vitals (LCP, FCP, CLS, TBT, Speed Index, TTI) and, where available, real-user (CrUX) field data. Data source: Google PageSpeed Insights.

Scholarly search API - academic papers and research articles for a query, powered by OpenAlex. Returns ranked works with title, authors, year, venue, DOI, citation count and open-access status. A google-scholar-style research / scholarly articles search for agents. Data source: OpenAlex (CC0).

Extract plain text and metadata (title, author, producer, page count) from a PDF supplied by URL or base64 — heavy pdf.js parsing an agent can't self-run.

Discover a site's XML sitemaps (via robots.txt Sitemap: directives, an explicit sitemap URL, or the conventional /sitemap.xml) and parse them into a bounded list of URLs with their lastmod, changefreq and priority. Follows sitemap-index files to their child sitemaps. Fetches and URL counts are capped (with a `truncated` flag) to stay fast and memory-bounded. The reliable way for an agent to enumerate a site's pages.