Data x402 offerings

Detect meaningful changes in normalized static page metadata and JSON-LD with a deterministic semantic hash. Transport noise and malformed JSON-LD counts are excluded; callers can fetch full details …

Deep-compare two caller-supplied JSON values with deterministic object-key ordering and explicit index-based array semantics. Returns bounded added, removed and changed paths with primitive before/af…

Extract one bounded value from caller-supplied JSON using a deterministic simple path such as items[2].name or $.data.id. Forest performs no network request or model inference, returns an explicit fo…

Time-series forecasting: forecast any BigQuery table with Google's pretrained TimesFM model (AI.FORECAST), nothing to train. Point it at the table, timestamp column and value column and get forecast rows with bounds for the horizon you set, per series with id_cols. Use it for sales, traffic, demand or metric forecasts. Input: table, timestamp_col, data_col; optional horizon, id_cols.

Dataset file to JSON rows: a CSV, TSV, XLSX or JSON link from any portal, an e-Stat (Japan) download or a data.go.kr (Korea) dataset page. Encoding detected (Korean, Japanese, Chinese, UTF and Western code pages), header row found, numbers typed, XLSX sheets selectable, large files capped and disclosed. Use it after opendata.search or on any data file URL. Input: url; optional max_rows, sheet, encoding, delimiter.

BigQuery SQL: run a read-only SQL query, including against Google's public datasets (Wikipedia, GitHub, blockchains, weather, census and more), and get columns and rows back as JSON with the bytes processed. Every query is dry-run first and refused above the scan ceiling, so cost is bounded before it runs. Input: sql; optional max_scan_gib. Ask in plain English instead: data.question.

Government open data search across 16 national portals in one call: Korea (data.go.kr), Japan, Hong Kong, Australia, UK, Canada, the EU catalog (139,000+ datasets), Germany, Italy, Ireland, Switzerland, Netherlands, Israel and Chile. Query in any language; each match returns organization, update time and download URLs ready for opendata.table. Portals that did not answer are listed. Input: query; optional region, limit, portals.

Ask a database in plain English (text to SQL): any BigQuery table, including Google's public datasets. The SQL is written for you, run under a byte ceiling, and the rows are read back into a written answer, returned with the SQL, columns and rows it used. Use it for analytics questions without writing SQL. Input: question, table; optional columns.

Anomaly detection on a time series: score recent periods against history with BigQuery AI.DETECT_ANOMALIES (TimesFM). Give one table and its latest periods are scored against its own history, or a history and a target table. Returns every checked row with bounds, anomaly flag and probability, plus the anomaly count. Input: history_table, target_table, timestamp_col, data_col; optional target_last, threshold, id_cols.

Structured data extraction / text to JSON: pull the fields you name out of any text into a JSON object with exactly those keys, null where the text does not state a value (Gemini, no invented values). Use it to turn emails, invoices, pages or documents into clean records. Input: text, fields[]. From a live web page: research.page_facts.

Dataset file to JSON rows: a CSV, TSV, XLSX or JSON link from any portal, an e-Stat (Japan) download or a data.go.kr (Korea) dataset page. Encoding detected (Korean, Japanese, Chinese, UTF and Wester…

Time-series forecasting: forecast any BigQuery table with Google's pretrained TimesFM model (AI.FORECAST), nothing to train. Point it at the table, timestamp column and value column and get forecast …

Structured data extraction / text to JSON: pull the fields you name out of any text into a JSON object with exactly those keys, null where the text does not state a value (Gemini, no invented values)…

Anomaly detection on a time series: score recent periods against history with BigQuery AI.DETECT_ANOMALIES (TimesFM). Give one table and its latest periods are scored against its own history, or a hi…

Find structured government datasets across Asia-Pacific open-data portals in one call: data.gov.au (Australia), the e-Gov Data Portal and the Tokyo catalog (Japan), DATA.GOV.HK (Hong Kong). Query in …

Predictive probability and regression engine: least-squares linear regression with standard errors, confidence and prediction intervals, a fitted normal distribution with quantiles and probability qu…

Use when an agent must decide whether a bank account it was given can be paid before trying: validates one IBAN offline - country, expected length, ISO 7064 mod-97 check digits, whether it is inside SEPA, and the account number split out. Deterministic, no network call and no third party that can go down. Built for cleaning a list of accounts one at a time.

Deterministic IBAN check before a SEPA/international transfer: format regex, per-country length (public SWIFT registry, ~85 countries) and ISO 7064 mod-97 checksum. Query: ?iban=FR1420041010050500013M02606 (spaces/dashes tolerated). Returns valid, country, bban, and a precise failure reason. Pure offline computation, 1y cache.

Parse iCalendar (.ics) text into structured JSON events: summary, start/end, location, organizer, attendees, status, and RRULE. Optional bounded recurrence expansion (expand:true, capped occurrences). Deterministic, pure CPU - send the ICS text, not a URL.

Extract text from a PNG or JPEG image - full text, overall confidence and per-line bounding boxes - from a URL or base64 payload, Tesseract on-device (no upstream API). Default English; other ISO 639-2 languages on request. Use it when an agent needs the words in a screenshot, scan or photo.

Classify up to 100 declared measurement intervals against tolerances as pass, fail or hold. Integer millimetres; uncertainty is an absolute interval half-width, not a confidence level. No sensor verification or engineering approval.

Classify up to 100 declared measurement intervals against tolerances as pass, fail or hold. Integer millimetres; uncertainty is an absolute interval half-width, not a confidence level. No sensor verification or engineering approval. New modular adapter: retained result recovery for 24 hours, minimal receipt metadata for 90 days. Existing live route unchanged.

Person name matching and deduplication: generates an AI-powered similarity key for an individual's full name, so variations such as 'Bob Smith', 'Robert Smith', and 'Smith, Robert J.' produce the same key. Group records on the key to find duplicate contacts and customers across CRM, marketing, and customer datasets.

City name standardization: converts abbreviations, nicknames, alternate spellings, and local variations ('San Fran', 'L.A.') to the standard city name. An optional state improves accuracy. For address cleansing, geocoding preparation, and consistent reporting.