The broader extraction schema: line items, per-field confidence metadata, and enabled feature envelopes. Start here to evaluate Taggun.
What Verbose Extraction Returns
Verbose endpoints use the broader extraction schema. Alongside the core totals, tax, date, and merchant fields, a verbose response can include product line items, per-field confidence metadata, helper arrays, detailed merchant fields, and enabled feature envelopes when those values are produced for the document — plus scan metadata such as trackingId and elapsed. The three endpoints below accept the same inputs as their simple counterparts; only the response depth differs.
When To Choose Verbose
- You need product line items — individual products, quantities, and amounts are extracted by verbose endpoints only. See Line Item Extraction.
- You use advanced features — fraud-detection signals and other enabled features return their results as envelopes on the verbose response.
- You are evaluating Taggun — verbose shows everything the platform can extract for your documents.
Only need totals, tax, date, and merchant? The simple form returns the compact core-field schema.
Testing your own receipts?Use the Upload File (verbose) playground to inspect the full extraction response for your own documents — no code required. Full guide: Test With Your Own Receipts.
Endpoints in This Category
| Endpoint | Method and path | Example response |
|---|---|---|
| Upload File (verbose) | POST /api/receipt/v1/verbose/file | Example JSON |
| URL (verbose) | POST /api/receipt/v1/verbose/url | Example JSON |
| Encoded File (verbose) | POST /api/receipt/v1/verbose/encoded | Example JSON |
Choosing an Input Method
- Upload File — send the document itself as
multipart/form-data. The most direct route when the file is on hand. - URL — point Taggun at a receipt hosted at a public HTTPS URL. Suits documents already in cloud storage or a CDN.
- Encoded File — send base64-encoded content in a JSON body. Suits queue pipelines and serverless functions where multipart uploads are awkward.
Accepted document types and size limits are defined once in Document Types & File Requirements.
Example Responses
Each endpoint above has a nested Example JSON page showing a real source receipt and the complete canonical verbose response it produces. The same validated JSON asset backs the endpoint's 200 response example, its example page, and Understanding Your Results — if they ever disagree, the canonical asset wins. Coordinate arrays are emptied in published examples for readability, and enabled-feature fields can vary with your account's settings.
Related Guides
- Understanding Your Results — read a verbose response field by field.
- Data Fields That Taggun Extracts — the full field reference.
- Line Item Extraction — product-level extraction in depth.
- Test With Your Own Receipts — evaluate with your real documents.
- Document Types & File Requirements — accepted formats and limits.
Parameters
Every endpoint in this category authenticates with your API key and shares the optional parameters below alongside its input fields.
| Parameter | Type | Where | Description |
|---|---|---|---|
apikey | header | required, all endpoints | Your API key. See Authentication. |
file | file | required on Upload File | Supported receipt or invoice file per Document Types & File Requirements. |
url | string | required on URL | HTTPS URL of the receipt, reachable by Taggun. HTTP, localhost, loopback, link-local, and private-network targets are rejected. |
headers | object | optional on URL | HTTP headers to forward when fetching a protected HTTPS resource. |
image, filename, contentType | string | required on Encoded File | Base64 content (image is a legacy field name — PDFs accepted), file name, and MIME type. |
extractLineItems | boolean | optional, default true | Set to false to skip product line-item extraction. See Line Item Extraction. |
extractTime | boolean | optional, default false | Set to true to return the time when found on the receipt; otherwise time is 12:00:00.000. |
language | string | optional | Language hint; leave empty for automatic detection. Common hints include en, es, fr, jp, he, iw, et, lv, lt, fi, el, zh, th, ar, kh — examples, not an allowlist. |
near | string | optional | A location (city, state, country) to help search for the merchant. |
ipAddress | string | optional, ≤100 | End user's IP for GeoIP lookup; can add a root location object to verbose responses (separate from the extracted merchant address). |
ignoreMerchantName | string | optional | A merchant name to ignore if detected on the file (e.g. a customer name mistaken for the merchant). |
refresh | boolean | optional, default false | Set to true to force reprocessing if the receipt is already stored. |
incognito | boolean | optional, default false | Per-request storage opt-out. See Security, Privacy & Data Residency. |
userId | string | optional, ≤50 | End-user identifier returned with matched receipts, when available, so cross-user reuse can be investigated. |
containerId | string | optional, ≤50 | Scopes duplicate & similarity checks to scans sharing the same value. Omit to compare account-wide. |
subAccountId | string | optional, ≤50 | Attribute usage to a client or subdivision of your account for reporting. Does not create a separate account. |
referenceId | string | optional, ≤50 | Caller-supplied identifier. See Duplicate & Similarity Detection. |