receipt-extraction
Extracts structured JSON from receipts and invoices with Australian GST/ABN validation, per-line tax codes, confidence scores, and rationale.
README
Receipt Extraction
AU GST/ABN-aware receipt & invoice extraction — applies documented AU entertainment/ITC tax-code
rules per line item, returning a tax_code, a confidence, and a one-line rationale for each, not
just OCR text. You supply the document; get validated structured JSON: vendor, line items, tax, totals —
with Australian GST and ABN checksum validation built in from day one. One API call, no hand-rolled
vision-LLM prompts, no re-deriving ABN/GST rules from the ATO's own documentation.
What it does
- Accepts a receipt/invoice image or PDF (base64), returns a clean, versioned
ReceiptJSON object. - Validates the vendor's ABN (11-digit modulus-89 checksum) and cross-checks GST across line items and totals, excluding non-creditable GST on entertainment-coded expenses per the documented rule set.
- Each line item carries its
tax_codeplus aconfidence(0–1) and a one-linerationalenaming the actual rule applied (e.g. "entertainment ITC denied, no exception context", or the incidental-to-travel exception that flips a restaurant-style meal back to claimable) — so you can check the call, not just trust it. - Works as an MCP server (
extract_receipttool) or a plain HTTPS API — usable from Claude, Cursor, or any HTTP client.
GST handling reflects publicly documented Australian tax law and is provided as software output, not tax advice.
Example (MCP tool call)
{
"tool": "extract_receipt",
"input": { "document_base64": "<base64 image or PDF>", "mime_type": "image/jpeg" }
}
{
"vendor": { "name": "Acme Pty Ltd", "abn": "51824753556", "abn_valid": true },
"date": "2026-07-14",
"line_items": [
{
"description": "Widget",
"qty": 2,
"unit_price": 12.5,
"gst": 2.5,
"tax_code": "taxable",
"confidence": 0.95,
"rationale": "Standard-rated retail good; 10% GST applies."
}
],
"totals": { "subtotal": 25.0, "gst": 2.5, "total": 27.5 },
"schema_version": "1.0"
}
Pricing
- Free — 20 documents/month, no card required.
- Pay-as-you-go — $0.02/document, no minimum.
- Pro — $19/month for 1,500 documents (~$0.013/doc effective), with priority support.
Getting a key
The MCP server is live at https://receipt-extraction.acjlabs.com/mcp (the alternate
https://acjlabs-receipt-extraction.acjlabs.workers.dev/mcp address reaches the same deployment and
keeps working, so existing configurations need no change). Free-tier keys
self-serve — POST /v1/signup with { "email": "you@example.com" } returns your key directly in the
response, good for 20 documents/month, no card required. Store it immediately — it is shown once and
cannot be recovered if lost (re-signup for a new one). Paid keys provision the same way via a
one-time claim link once Pro/pay-as-you-go billing is live. See
https://acjlabs-receipts.pages.dev for the same steps plus pricing and a comparison against
alternatives.
Connecting it to an MCP client
With Claude Code:
claude mcp add --transport http receipt-extraction \
https://receipt-extraction.acjlabs.com/mcp \
--header "Authorization: Bearer YOUR_API_KEY"
Any other MCP client that supports a remote HTTP server with a custom header (Cursor, Cline, VS Code,
etc.) works the same way: point it at the URL above with an Authorization: Bearer YOUR_API_KEY
header. A gateway or scanner that can only forward the raw key value (no Bearer prefix) also works —
both forms authenticate. Tool discovery (initialize/tools/list) doesn't require a key at all; only
calling extract_receipt does.
See docs/quickstart.md for a full copy-paste walkthrough (getting a key, per-client
configs, confirming the connection with curl) if you'd rather follow one linear guide.
How accurate is it?
Validated 4/4 against hand-built (but genuinely readable) test documents covering net-priced, GST-inclusive-only, checksum-invalid-ABN, and unreadable-document cases. Real production accuracy across more document formats will be tracked once live traffic exists.
npm client
Prefer a typed function over hand-rolling MCP JSON-RPC calls?
@acjlabs/receipt-extraction-client
(live on npm) wraps the extract_receipt tool call:
npm install @acjlabs/receipt-extraction-client
import { createReceiptExtractionClient } from "@acjlabs/receipt-extraction-client";
const client = createReceiptExtractionClient({
baseUrl: "https://receipt-extraction.acjlabs.com",
apiKey: "...",
});
const receipt = await client.extractReceipt(base64EncodedImageOrPdf, "image/png");
Why not just use generic OCR?
Generic OCR extraction APIs return raw fields. None of them apply the AU layer on top: that an Australian ABN carries an 11-digit modulus-89 checksum, or that GST apportionment differs for entertainment-coded expenses. You'd build and maintain that yourself, against someone else's schema that can change under you. See the full comparison, or read why AU GST rules break most receipt-extraction tools (the ABN checksum, the three real GST cases, and how to test any tool against them).
Source availability & support
This repository hosts the documentation for the hosted service. The service implementation is not open source. Bug reports and feature requests are welcome in this repo's Issues; you can also reach us at contact@acjlabs.com.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.