Whether a site publishes /llms.txt and /llms-full.txt: the parsed llms.txt (title, summary, sections with their links) and the size of llms-full.txt. Reads the site itself, through the same URL guard and size cap as the page tools.

Items of an RSS, Atom or RDF feed, or of the feed a web page advertises: title, link, published date, author and a 300-character summary, newest first (max 50). since keeps items published on or after a date.

Sentiment of one text, or up to 20 (texts, a JSON array), each up to 2,000 characters: POSITIVE or NEGATIVE with the model's confidence, from DistilBERT SST-2. Best for short English text. It measures the wording, not whether the statement is true.

Public holidays of a country (ISO 2 code) for a year (default: this year): date, local and English names, and whether the holiday is nationwide or regional with its counties. Source: Nager.Date. An empty list is a valid answer.

Validates JSON data against a JSON Schema (draft 2020-12): valid true or false, and up to 50 errors with the instance path, the keyword and the message. Schema and data are each limited to 200 KB. Remote $ref targets are not fetched. No network call.

The current time in an IANA time zone (its UTC offset and whether DST applies) and, with time and to, the conversion of a local time from one zone to another. Computed in the Worker with the runtime's time zone data; no outside call.

Active US National Weather Service alerts for a state or a point: event, severity, urgency, area, onset, end of the event (ends), expiry of this message (expires), headline and instruction (cut to 400 characters). Attribution: National Weather Service, public domain.

A YouTube video's captions as plain text and timestamped segments, with the video's title, channel, duration, views, publish date, chapters and the list of caption languages. Accepts watch, youtu.be, shorts, embed and live links or the 11-character id; pick a language with lang (default English, else the first track). Auto-generated captions contain recognition errors. A video without captions is a 404 and not charged.

Structured data from a CV or resume: name, contact details, location, links, summary, skills, languages, certifications, work experience and education, with null or empty lists where the text has nothing. Plain text, or a PDF with a text layer by URL or base64. Dates are kept as written. Not live-checked: Workers AI is checked by the owner after the merge.

Structured data from an invoice: seller and buyer (name, address, tax ID), invoice number, dates, currency, line items, subtotal, tax, total and payment details, with null for anything absent. The arithmetic is checked (line sum against subtotal, subtotal plus tax against total) and mismatches are returned. PDF with a text layer by URL or base64, or plain text. Not live-checked: Workers AI is checked by the owner after the merge.

Geocodes a one-line US address with the US Census Bureau: matched address, coordinates, county, state, census tract and block group. Addresses that do not match answer 404 (not charged). Public domain data.

Weather for a place or a coordinate: current conditions, the next 48 hours and 7 days (temperature, precipitation, wind, symbol). Forecast from MET Norway, cached until its own expiry time. Attribution: Data from MET Norway (CC BY 4.0).

Translate an SRT or WebVTT subtitle file into another language. Only the cue text is translated; every cue number, timing line and header is copied exactly, and the output is checked to parse back to the same cues and timings, or it is refused unpaid. Body { subtitles, to, from }: up to 50 KB. Workers AI (Llama 3.1 8B) translates the text, so check it before publishing. Not live-checked: the owner tests it on the Worker after the merge.

A security by ISIN, CUSIP or ticker (with an optional exchange code): its OpenFIGI id, name, ticker, exchange, security type and market sector, and for an ISIN the issuer's LEI and legal name from GLEIF. ISIN and CUSIP check digits are verified before the call. Attribution: FIGI by OpenFIGI; LEI by GLEIF.

Screen one person, company, vessel or aircraft name against the OFAC SDN and Consolidated lists, the EU consolidated list, the UN Security Council list and the UK Sanctions List (official files, refreshed daily). Fuzzy matching with aliases, transliteration and an optional country and date of birth to sharpen the score. Returns each possible match with its list, entry ID, score, programmes and official link. A screening aid, not a legal determination.

Google PageSpeed Insights for one public URL: the Lighthouse scores (performance, seo, accessibility, best-practices), the lab Core Web Vitals (FCP, LCP, CLS, TBT, Speed Index), the CrUX field data when Google has it (LCP, INP, CLS), and the top five opportunities with estimated savings. Mobile by default. A run takes 10 to 40 seconds and is cached for an hour. Keyless calls share Google's quota, so a busy moment is a 503.

Perpetual futures funding for one coin across Hyperliquid, OKX, Deribit and dYdX v4: the current funding rate, the interval (1 or 8 hours), the annualised rate, the next funding time, open interest in coins and USD, and the mark or oracle price. A venue that does not list the coin is left out and named in unavailable. Rates are as the venues publish them; funding is paid between traders, not to the venue. Information only, not financial advice.

Subdomains of a domain seen in certificate transparency logs: each name with the first and last certificate dates, the issuer of the newest certificate, and wildcard entries flagged. Deduplicated, newest first, up to 500. CertSpotter is read first, but its free index is partial, so those answers are marked truncated with the reason; crt.sh is the fallback and is complete. Only public certificates are listed.

Points of interest near a point (lat and lon, or a place name geocoded through OpenStreetMap) within up to 5 km: pharmacies, restaurants, cafes, hospitals, ATMs, banks, fuel, EV charging, hotels, supermarkets, parking, schools, police, toilets. Name, distance, address parts, opening hours, phone and website when tagged. OpenStreetMap data (ODbL). Up to 50 results.

Packaged food by barcode (EAN-8, UPC-A, EAN-13 or GTIN-14) or by name: product name, brand, quantity, Nutri-Score, NOVA group, nutrients per 100 g (energy, fat, saturated fat, carbohydrates, sugars, fibre, protein, salt), allergens, labels, ingredients text and front image URL. Open Food Facts data (ODbL, attribution in source). A barcode not in the database is a 404, not charged. Community data: values can be missing or wrong.