Receipt / document recognition and archiving

Disclaimer: The content shown on this page is generated by an AI model and is for reference only. It does not constitute any guarantee of effectiveness. Actual output and Credits consumption are subject to "Usage Details" in the console.

This page runs through a real invoice (sensitive fields redacted): hand the receipt image to qwen3.7-plus and in one sentence get structured JSON plus a suggested file name generated by the rule "date_seller_amount"—image reading, field extraction, and named archiving in one shot. Call chain: qwen3.7-plus (visual understanding, per-image OCR + structuring) → rename and archive by field.

Measured output

Run through with a real invoice (sensitive fields redacted). The OCR structured output:

{"Type": "Electronic invoice (general invoice)", "Issue date": "January 20, 2026", "Seller": "Redacted", "Amount": "28.87", "Tax": "1.73", "Total (price + tax)": "30.60", "Invoice number": "Redacted"}
2026-01-20_Redacted_28.87

Steps: Invoice recognition and archiving

Send the receipt image along with the following instruction to qwen3.7-plus:

Recognize this invoice and extract it as JSON: {"Type":"","Issue date":"","Seller":"",
"Amount":"","Tax":"","Total (price + tax)":"","Invoice number":""}.
Fields obscured by gray blocks in the image (redacted) should have the value "Redacted"; do not reconstruct from visible fragments.
Then give a suggested file name in the format "date_seller_amount". Output only the JSON and the file name on two lines.

Notes

Consumption: This playbook involves visual understanding (qwen3.7-plus reading the image).
Boundaries: The image itself is converted into tokens and counted toward input; this is the main consumption source for this type of task. The larger the image, the more tokens it typically consumes.
Redaction: The company name, tax ID, invoice number, and issuer in the sample invoice have all been redacted.

​Measured output

​Steps: Invoice recognition and archiving

​Notes

Measured output

Steps: Invoice recognition and archiving

Notes