How to Convert PDF Invoices to Xero: 6 Methods Compared

Xero does not natively extract data from PDF invoices. Six methods exist for getting that data in: manual rekey, email-to-bills, Hubdoc, third-party OCR tools, AI extraction to CSV import, and AI extraction with direct API push. Only the last two support full line item extraction with batch processing and multi-currency handling.

This guide compares all six methods for converting PDF invoices to Xero side by side, evaluating each on line item support, batch capacity, multi-currency handling, and cost so you can identify exactly which approach fits your volume and complexity.

---

Why Xero Has No Built-In PDF Invoice Extraction

Xero does not read PDF invoices. There is no upload button, no background parser, and no planned feature that will extract supplier details, amounts, or line items from a PDF and populate a bill automatically. Xero has confirmed it has "no plans to extract data from a PDF to automatically create an invoice" natively.

The closest built-in option is the Xero Conversion Toolbox, but its name is misleading. The Conversion Toolbox accepts CSV files only and exists for a specific purpose: migrating historical data from other accounting systems (MYOB, QuickBooks, Sage) into Xero. It maps columns from a structured CSV into Xero's chart of accounts, contacts, and opening balances. It was never designed for ongoing invoice processing and cannot handle PDFs in any form.

This leaves a significant workflow gap for anyone who needs to import PDF invoices into Xero on a recurring basis. Your suppliers send structured data (amounts, dates, line items, tax codes) locked inside PDF files, but every Xero import path requires that data in a structured, machine-readable format like CSV before it will accept it. The extraction and reformatting step falls entirely on you, and for practices processing invoices regularly, it is a tedious manual bottleneck.

Xero has announced AI-powered features under the JAX brand, with initial rollout beginning in the UK. However, these are not yet widely available, and existing Xero users still face the same PDF extraction gap. Six distinct methods currently exist to bridge it, each with different trade-offs in accuracy, automation level, and line item support.

---

All Six Methods at a Glance

The following Xero data capture tools comparison lays out all six methods side by side across the capabilities that matter most: whether line items are extracted, accuracy level, batch processing capacity, multi-currency support, and typical cost model.

Method	Line Items	Accuracy	Batch Size	Multi-Currency	Cost Model
Manual rekey	Yes (manual)	High (if careful)	N/A	Yes (manual)	Staff time
Email-to-bills	No (attaches PDF only)	N/A	One at a time	No	Free (built into Xero)
Hubdoc	No (header-level only)	Moderate	Limited	No	Free with Xero subscription
Third-party OCR (Dext, AutoEntry, etc.)	Varies by tool and plan	Moderate to high	Moderate	Varies	Subscription ($20-70+/month)
AI extraction to CSV import	Yes	High	Thousands of files	No (CSV limitation)	Pay-per-page
AI extraction with API push	Yes	High	Thousands of files	Yes	Pay-per-page

Methods 1 through 3 are the most common starting points for Xero users, while method 4 adds dedicated extraction tools. Methods 5 and 6 represent a newer category that directly addresses the line item extraction and batch processing gaps in the traditional options.

---

Manual Rekey, Email-to-Bills, and Hubdoc

These are the three methods most Xero users try first. Each has a clear use case, but all three share the same critical gap.

Method 1: Manual Data Entry

The most straightforward way to convert PDF invoices to Xero is the most tedious: open the PDF, read each field, and type it into Xero's bill entry screen. Supplier name, invoice number, date, due date, account codes, tax rates, line descriptions, quantities, unit prices, every value entered by hand.

For a handful of invoices per week, manual rekey is manageable. At higher volumes, two problems compound. First, data entry errors become inevitable. Transposed digits, missed line items, and incorrect tax codes creep in as attention drifts across dozens of invoices. Second, the staff time cost scales linearly: ten times the invoices means ten times the keystrokes. There is no efficiency gain at volume.

Manual entry is the baseline. Every other method on this list exists to reduce or eliminate it.

Method 2: Email-to-Bills

Xero assigns each organisation a unique email address (found in your Xero organisation settings). Forward a PDF invoice to that address, and Xero creates a draft bill with the PDF attached. The workflow sounds promising until you open the draft.

Xero does not extract any data from the PDF. The draft bill arrives with blank fields. You still need to manually enter the supplier, date, amounts, account codes, and every line item. The email-to-bills feature is essentially a filing system: its primary value is attaching the source PDF to the bill record so you can reference the original document later.

For users who assumed emailing invoices to Xero would scan and populate bill data automatically, the email-to-bills limitations are a common frustration. The attachment convenience is real, but the data entry workload remains unchanged.

Method 3: Hubdoc

Hubdoc comes bundled with Xero subscriptions at no additional cost, making it the default next step for users who want to scan invoices to Xero with some degree of automation. It can fetch documents from suppliers automatically and extract certain data fields.

The extraction, however, is limited to header-level information: supplier name, invoice date, invoice number, and total amount. Hubdoc does not extract individual line items. If your invoices contain multiple lines with different account codes, quantities, or descriptions, you will need to add those manually in Xero after Hubdoc pushes the header data across.

Users have also reported GST rounding issues when Hubdoc calculates tax, which can require manual correction before reconciliation. On the Xero App Store, Hubdoc holds a 3.5 out of 5 rating, reflecting a tool that works adequately for straightforward single-currency invoices with simple headers but struggles with anything more detailed.

The common thread across all three methods: none extract line item data from PDF invoices. For practices that need full line items in Xero, with correct account codes, tax rates, and descriptions per line, dedicated third-party tools offer deeper extraction capabilities.

---

Third-Party OCR and Data Capture Tools

Beyond Xero's native options, a category of dedicated data capture tools connects directly to Xero through its API. These tools go further than Hubdoc by using OCR (optical character recognition) combined with rules-based extraction to read PDF invoices and create bills in your Xero account automatically.

Four tools dominate this space for Xero users:

Dext (formerly Receipt Bank) is the most widely adopted, carrying a 4.8 out of 5 rating on the Xero App Store. It operates on a subscription model with tiered plans, and line item extraction is available only on higher-tier subscriptions. For firms processing invoices across multiple clients, Dext offers bulk upload and a fetch feature that pulls documents from email inboxes and cloud storage.

AutoEntry holds a 4.7 out of 5 Xero App Store rating and relies heavily on rule-based matching. Once you process an invoice from a given supplier, AutoEntry applies the same extraction rules to future invoices from that vendor. The subscription pricing scales with document volume, which can become costly as your processing needs grow.

Datamolino scores 4.9 out of 5 on the Xero App Store and differentiates itself with a review workflow. Extracted data is presented for human verification before being pushed to Xero, giving you a correction step that reduces errors in your accounting records.

EzzyBills rounds out the Xero marketplace options as a scanning tool built specifically for invoice processing, with direct Xero integration for creating bills from uploaded PDFs.

These tools share several trade-offs worth weighing. Monthly subscription fees apply regardless of whether you process five invoices or five hundred in a given period. Line item extraction quality varies, with some tools restricting it to premium tiers or struggling with complex table layouts. Because they rely on templates and rules, they can misread invoices from new suppliers until you manually correct and retrain the extraction logic. For example, if a supplier changes their invoice layout or uses an unfamiliar table structure, Dext or AutoEntry may map line items to the wrong fields until you manually fix and save a new extraction rule. Batch processing capacity also has limits compared to newer approaches.

If you are comparing tools in this category, a structured framework helps cut through marketing claims. Our guide on evaluating invoice scanning software walks through the criteria that matter most for accuracy, cost, and integration depth.

---

AI-Powered Extraction: CSV Import and Direct API Push

The first four methods share a common ceiling: limited line item support, low batch capacity, or no multi-currency handling. AI-powered extraction removes those constraints entirely. Two approaches exist here, and the difference between them is whether you use Xero's CSV import as an intermediary or bypass it altogether with a direct API push.

Method 5: AI Extraction to CSV Import

The workflow is straightforward. An AI-powered invoice data extraction tool reads your PDF invoices, extracts every field you need, including full line item detail, and produces a structured CSV or Excel file formatted to match Xero's Conversion Toolbox import template. You then upload that CSV into Xero, where it creates draft bills with line items intact.

Invoice Data Extraction handles this at scale. You can upload up to 6,000 mixed-format files (PDF, JPG, PNG) per job, or single PDFs up to 5,000 pages. Processing runs at 1-8 seconds per page, with speeds often reaching 2 seconds per page on larger batches. The key advantage over traditional OCR tools is prompt-based control: you write natural language instructions telling the AI exactly what to extract and how to structure the output. A prompt like "Extract invoice number, date, vendor name, line items with description, quantity, unit price, and line total, formatted for Xero CSV import" gives you a file ready to upload without manual column mapping.

The platform outputs in Excel (.xlsx), CSV (.csv), or JSON (.json), and supports all major languages and scripts, which matters if you process invoices from international suppliers. Users whose primary goal is Excel output rather than Xero import can find a dedicated guide on converting PDF invoices to Excel.

The limitation with method 5 is on Xero's side, not the extraction side. Xero's CSV import cannot handle multi-currency invoices. The Conversion Toolbox template also does not match Xero's own export format, so you cannot round-trip data between export and import. For a detailed walkthrough of the CSV upload process itself, see our step-by-step guide to importing invoices into Xero via CSV.

Method 6: AI Extraction with Direct API Push

This method removes the CSV intermediary entirely. The AI extraction tool reads your PDF invoices and pushes structured data directly into Xero via the Xero API, creating bills with full line item detail in one automated step.

The extraction side works identically to method 5: the same batch capacity, the same processing speed, the same prompt-based control over what gets extracted. The platform's RESTful API accepts PDFs and images, processes them with your natural language instructions, and returns structured JSON results. From there, an integration layer maps the extracted data to Xero's API schema and creates bills programmatically.

Method 6 solves the three limitations of CSV import:

• Multi-currency support. The Xero API accepts currency fields that the CSV template cannot accommodate. International invoices in EUR, USD, AUD, or any other currency flow through correctly.
• No manual file handling. There is no CSV to download, format-check, and upload. The data moves from PDF to Xero bill without human intervention.
• Direct bill creation. Bills appear in Xero as drafts with full line items, ready for review and approval.

For users evaluating whether an API-based approach, a SaaS platform, or an ERP-embedded tool is the right architecture for their practice, we have a guide on choosing between API, SaaS, and ERP invoice capture that covers the trade-offs in detail.

The right method between these two depends on your invoice volume, the complexity of your supplier base, and whether you need line items and multi-currency support.

---

Choosing the Right Method for Your Invoice Volume

The right method depends on three variables: how many invoices you process per week, whether you need line item detail, and whether you deal with multiple currencies. Here is a practical framework to match your situation to the best approach.

Under ~20 invoices per week, single currency, header-level data only. Manual rekey or email-to-bills handles this volume without significant overhead. If you want to reduce data entry slightly, Hubdoc captures header-level fields like supplier name, date, and total amount, but it will not extract individual line items. At this scale, the cost of a dedicated tool rarely justifies itself.

20 to 100 invoices per week, some line items needed, single currency. Third-party OCR tools such as Dext, AutoEntry, or Datamolino deliver meaningful time savings here. Each platform differs in pricing, line item accuracy, and Xero integration depth, so evaluate based on which capabilities matter most to your workflow and what fits your subscription budget. This tier is where most small-to-mid-size practices find the clearest return on investment.

100+ invoices per week, full line items required, multi-currency, or batch processing. AI extraction with CSV import or direct API push is the only category that satisfies all of these requirements simultaneously. When you are processing invoices across multiple suppliers, currencies, and tax jurisdictions at volume, traditional OCR tools hit accuracy and throughput limits that AI-powered extraction does not.

The financial case for moving up tiers is well documented. According to CPA Australia's 2024 Business Technology Report, which surveyed over 1,000 accounting and finance professionals across Asia-Pacific, 76 per cent of businesses using AI reported that their profitability increased in the last financial year, compared to just 34 per cent of businesses that did not use AI. For accounting practices serving multiple Xero clients, the transition from manual methods to AI-assisted extraction delivers measurable financial impact that compounds across every client engagement.

Three steps you can take today:

1. Identify your tier. Count your weekly invoice volume and note whether you need line item extraction or multi-currency support. This alone narrows the six methods down to one or two realistic options.
2. Test before you commit. For AI extraction methods (CSV import or API push), run a small batch of 10 to 20 real invoices through the tool before changing your production workflow. Verify that supplier names, line items, tax codes, and totals extract correctly against your actual documents.
3. Use the right import format. If you choose the CSV import route, structure your output to match the Xero Conversion Toolbox import template. This ensures your extracted data maps cleanly into Xero without manual column remapping or failed imports.

Xero invoice automation does not require a single leap from manual entry to full AI extraction. Start at the tier that matches your current workload, validate the accuracy with real invoices, and scale up as your volume or complexity grows.