General Utilities x402 offerings

Hybrid vs gas savings calculator. Compares annual fuel cost between a hybrid and a conventional gas vehicle based on miles driven, gas price, and each vehicle's MPG, and computes the break-even point for the hybrid's price premium.

UUID generator API. Generates a random RFC 4122 version 4 UUID using a cryptographically secure source — not the pseudo-random, collision-prone IDs an LLM would invent. No input required.

Checksummed-financial-identifier generator API. Generates a structurally valid iban, bic, creditCardNumber, bitcoinAddress, litecoinAddress, ethereumAddress, or routingNumber, each computed with its real checksum algorithm (mod-97, Luhn, EIP-55, etc.). Accepts optional locale (default en_US).

Date and time calculations in pure code. op=diff: calendar years, months and days between two dates, and the total days. op=add or subtract: move a date by years, months, weeks, days, hours or minutes (month ends are clamped). op=info: weekday, ISO week, day of year, quarter, and age from birthDate. op=tz: a local date-time in one IANA time zone shown in another. Impossible dates are a 400 with the reason.

Parses a User-Agent string: browser and version, rendering engine, operating system and version, device type, vendor and model, CPU, and whether it is a known crawler (Googlebot, Bingbot, GPTBot, ClaudeBot, PerplexityBot, AhrefsBot and a few more) with its name. Optional Client Hints values (platform, mobile, model) are reported as given. Parsed with ua-parser-js 1.x (MIT). Pure code, no outside request.

A QR code for a text of up to 300 characters (a URL, Wi-Fi login or contact card), generated locally. format=svg returns the image itself as image/svg+xml; format=json returns the version, the module count, the SVG and a data URI. Error correction L, M, Q or H (default M); size from 128 to 1024 pixels (default 512).

Labels up to 20 texts with 2 to 20 labels you choose, with a confidence from 0 to 1 and a one-line reason per text. One label per text, or several with multi true. Uses the chat model with JSON output; labels are as good as the model's reading of them.

Camera make and model, lens, date taken, exposure (ISO, shutter, aperture, focal length), orientation, image size, software, and GPS latitude, longitude and altitude with a map link, from the EXIF of a JPEG, HEIC, TIFF or PNG image at a URL. Only the first 256 KB is read. A file that is not an image is 415; an image without EXIF is 200 with hasExif false.

Explains a cron expression in plain English and lists its next runs in UTC and in a time zone. Five fields (minute hour day month weekday), or six with seconds first. Names (MON, JAN) and steps are supported. Computed in the Worker; no network call.

Split a long text into chunks for retrieval (RAG): split on paragraphs, then sentences, then words, each chunk within maxTokens (default 500) with the overlap you ask for (default 50 tokens). Each chunk has its index, text, start and end character offsets and an estimated token count (characters divided by 4, a rough rule for English). Up to 1 MB and 2,000 chunks.

Checks the check digit and structure of an identifier with no network call: ISBN-10 and ISBN-13 (with ISBN-10 to 13 conversion), EAN-8, UPC-A, EAN-13 and GTIN-14 (with the GS1 prefix country), VIN (manufacturer code, region, model year), IMEI, ISIN (country) and a generic Luhn number. type=auto detects the type. Up to 50 comma-separated values give a results array.

Working days for a country. mode=between counts the working days after `from` up to and including `to` (up to 5 years) and lists the public holidays skipped; mode=add gives the date that is N working days after `date` (up to 3,650). Weekends are Saturday and Sunday, Friday and Saturday in Israel, and Friday only in Iran and Afghanistan. Nationwide public holidays from Nager.Date; a country with no holiday data is a 404, not charged.

Reads the text of an image (JPEG, PNG, GIF or WEBP, up to 4 MB) with Llama 3.2 11B Vision on Workers AI. Send an https url or base64 image, and optionally a language hint. Returns the full text with its line breaks, the lines, a language guess and the model that answered. An image with no readable text is a 422 and is not charged.

Transliterate text to Latin ASCII, or make a URL slug from it. Covers Cyrillic, Greek, Turkish, German, Nordic, Polish, Vietnamese and accented Latin letters. Characters with no table entry (Arabic, Hebrew, CJK ...) are kept and listed in unconverted, not guessed. Up to 2,000 characters. Pure code.

Translation with Llama 3.1 8B: up to 6,000 characters per call, formatting, names, numbers and URLs kept, source language detected when not given. Language codes like en, tr, de, fr, es, ar, zh, or language names.

Answers a question from the given text only (max 30,000 characters), with supporting quotes and their character offsets. Each quote is checked against the text and dropped if it does not occur in it. answerable is false when the text does not contain the answer. Model output: verify before relying on it.

Differences between two texts: mode lines, words or chars; the added, removed and unchanged parts with counts, a similarity ratio (2 x unchanged / both texts' units), and a unified diff for lines. Body { a, b, mode }: lines and words up to 50 KB each, chars up to 20 KB each (capped for the Worker's CPU budget). Pure code using the diff package (BSD-3-Clause).

The language of a text with franc-min: ISO 639-3 and ISO 639-1 codes, the English name, confidence from 0 to 1, the script (Latin, Cyrillic, Arabic...) and the three next most likely languages. Send texts= (a JSON array of up to 20 strings) for a batch. Under 20 characters the answer is und with a note. 61 languages.

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.

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.

Encode or decode text as base64, base64url, hex, URL (percent) encoding, HTML entities, punycode (domain names) or quoted-printable. Body { input, scheme, direction }, up to 1 MB. Invalid input to decode is a 400 that names the position of the first bad character. Unknown HTML entities are reported, not guessed. Pure code, no outside request.

Checks and formats a phone number with libphonenumber: valid and possible flags, E.164, international and national formats, country, calling code and line type (mobile, fixed line, toll free and so on). Without a + sign, give the country. No network call.

Multilingual text embeddings (BGE-M3, 1024 dimensions) for up to 100 texts per call. For search, RAG and clustering.

Convert a value between units of length, area, volume (US customary and metric, cooking measures), mass, speed, pressure, energy, power, data size (SI and binary), time and temperature. Factors are exact to the definitions in NIST SP 811 (international inch and pound, US gallon, calorie IT, Btu IT, mechanical horsepower). An unknown unit is a 400 that lists every supported unit. Pure code, no outside source.