Interactive Browser Demo
Try pacs008 in Your Browser
Drop in a CSV, use the sample, or paste your own records — payment payloads are processed locally by your browser and are not transmitted to pacs008.com. Follow your organization's data-handling policy and use synthetic or masked test data where required.
Add your payment data
Drag a CSV file here — or click to browse
Maximum 2 MB · read locally by your browser, never uploaded
Validation result & generated pacs.008 XML
Add data in step 1 — validation runs automatically.
| Row | Column | Rule | Value | Message |
|---|
Inspect an existing XML file
Check a pacs.008 file you already have, rather than generating a new one. Parsing happens in this page; nothing is uploaded.
Validate against the XSD
Checks element order, cardinality and datatypes against the official ISO 20022 schemas, using libxml2 compiled to WebAssembly. The message type is detected from the document namespace. Schemas are published here for pacs.008, pain.001, pain.002, pain.007, pain.008, camt.110 and camt.111 — see the Trust Centre for exact versions. Anything else reports not evaluated. The engine is about 800 KB and is downloaded only when you press the button, then cached.
XSD covers structure only. Scheme rules such as CBPR+ address requirements are checked separately above, and ISO semantic consistency is still not evaluated in the browser — see the Trust Centre. If the engine or schema cannot be loaded, the result says not evaluated rather than reporting a pass.
Batch address readiness scan
Upload a CSV of party addresses to see how many records would still be accepted after 14 November 2026, when fully unstructured postal addresses are removed from the applicable CBPR+ and CHAPS flows.
The file is read in this page and classified in memory. It is not uploaded, not stored, and not logged. Use masked or synthetic data if your policy requires it.
Recognised columns: street, building_no, town, country, address_line1, address_line2, party, name. Names are matched loosely, so TwnNm and town_name both work.
Classification uses the same rules as the address checker above, so a batch result and a single record cannot disagree.