Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

invoice2data

github.com

  • Free plan
  • Windows
  • macOS
  • Linux
  • Self-hosted
7.6 Club score Visit github.com

What invoice2data is

invoice2data is a command-line tool and Python library that extracts structured information from PDF invoices for accounting workflows. It reads PDF text through selectable backends, including pdfium, pdftotext, pdfminer, pdfplumber, and OCR tools, then matches that text against YAML or JSON templates. Templates can specify custom fields and multiple regular expressions per field. The command-line tool can export results as CSV, JSON, or XML, and rename PDF files to reflect their contents. Optional plugins support invoice line items and tables, with Camelot available for advanced table parsing. Optional AI features can help draft templates from sample documents or extract fields when template matching fails. AI does not run unless a provider is configured; the default mock provider has no network access. Listed AI providers include OpenAI, DeepSeek, Mistral, Gemini, and Ollama, which can run locally. The project also describes docTR and PaddleOCR as local OCR options. It is free under the MIT License, with pip and conda installation routes. Its README defers canonical schemas for non-invoice documents such as waybills and purchase orders.

Who it suits

invoice2data suits developers and accounting workflows that need invoice data extracted from PDFs using templates. It is also relevant where local OCR or optional AI extraction is useful.

Good to know, and what to watch

  • Command-line and Python library interfaces
  • Exports to CSV, JSON, or XML
  • Templates support custom fields and multiple regexes
  • Optional plugins handle line items and tables
  • AI is inactive unless a provider is configured
  • Canonical schemas for non-invoice documents are deferred
  • AI extraction requires configuring a provider

invoice2data

Free

Open source

The Club's review

Overview

invoice2data is a free, open-source command-line tool and Python library for turning PDF invoices into structured data for accounting workflows. It suits developers, IT teams and organisations prepared to configure extraction rules and run software themselves. The trade-off is clear: you get control and local deployment rather than a packaged invoice-processing service.

Key features

Text extraction can use pdfium, pdftotext, pdfminer or pdfplumber, with OCR tools available for scanned documents. This choice helps teams adapt the parser to their PDF mix, while OCR backends require their own setup and tuning. The installation documentation says the default pdfium route needs no system libraries on Windows, macOS or Linux, reducing the initial platform friction.

YAML or JSON templates describe the fields to capture and can assign several regular expressions to each field. That is a strong fit for recurring supplier layouts and unusual fields, but the responsibility for creating, testing and maintaining those rules remains with the user. Changes to invoice designs therefore become an ongoing engineering task rather than a vendor-managed configuration.

The command-line interface exports CSV, JSON or XML and can rename PDF files according to their contents. Optional plugins add line-item and table matching, including an opt-in Camelot plugin for more advanced table parsing. These capabilities make it suitable for accounting pipelines that need item-level records, provided the team is willing to select and configure the relevant plugins.

Optional AI functions can draft templates from sample invoices or act as a fallback when templates do not match. The listed providers are OpenAI, DeepSeek, Mistral, Gemini and Ollama. AI remains inactive unless a provider is configured, and the default provider is a network-free mock. Ollama can run locally; the documented docTR and PaddleOCR OCR backends also keep processing on the user's machine. That gives privacy-conscious deployments a local path, while cloud providers introduce the usual need to assess data handling before enabling them.

The project is used by an Odoo OCA module, offering a concrete integration route for Odoo-based accounting environments. Its stated scope remains invoices: canonical schemas for waybills, delivery notes and purchase orders are deferred, so it is not a general document-understanding layer.

Pricing

invoice2data costs nothing, with an open-source licence and no paid tiers or usage limits stated. There are no seat charges or metered quotas in the published plan. The MIT License permits use, modification, distribution, sublicensing and sale subject to its conditions, which is useful when embedding the library in an internal or commercial system. Budgeting still needs to account for your own hosting, configuration and maintenance, because the project does not describe a paid support plan.

Platforms

It runs on Linux, macOS and Windows, and supports self-hosted deployment. Pip and conda installation are documented. Self-hosting suits organisations that need control over files and processing locations, while teams wanting a vendor-operated web service will need another category of product.

Who it's for

Choose invoice2data when you have developers or technically capable administrators, repeatable invoice formats and a requirement to control extraction logic or keep OCR and AI local. It is also a sensible building block for accounting integrations, including Odoo workflows, where CSV, JSON or XML outputs can feed downstream systems.

Look elsewhere if finance users need a ready-to-use interface, managed supplier onboarding, broad document schemas or vendor-backed support. The project invites issues and pull requests and links to a contributor guide, but no paid support offering is described. GitHub's security page reports no SECURITY.md policy and no published security advisories, so organisations with formal security-governance requirements should review that gap before adoption.

Pros and cons

Pros

  • Free and permissively licensed: the MIT License allows modification and commercial distribution without subscription fees.
  • Configurable extraction: selectable text and OCR backends plus YAML or JSON templates let technical teams tailor fields and regular expressions to their suppliers.
  • Structured outputs: CSV, JSON and XML exports, PDF renaming and optional line-item/table plugins support downstream accounting automation.
  • Local privacy options: self-hosting, local OCR and Ollama avoid sending documents off-machine when configured that way.

Cons

  • Template maintenance is your job: invoice-layout changes can require new or revised rules and testing.
  • Not a broad document platform: canonical handling for waybills, delivery notes and purchase orders is outside the stated scope.
  • Operational responsibility: there is no published paid support plan, and no SECURITY.md policy or advisories are published.
  • Optional components add complexity: table plugins, OCR engines and AI providers each require deliberate selection and configuration.

Alternatives

For a wider shortlist, browse Invoice OCR Software and Invoice Processing Software.

Doxis AI.dp is a paid, web/API and mobile option with free credits and usage-based billing, better suited to buyers wanting a hosted service. Veryfi Invoices OCR API fits teams seeking a managed API; its Starter plan is $500 per month with a stated minimum commitment and per-invoice pricing. Nanonets Invoice OCR offers cloud or self-hosted API access with a free starting allowance and paid credits, making it preferable when a vendor-managed AI workflow matters. Parseur is a freemium web/API parser with a free tier capped at 20 pages per month and one user, useful for small document volumes. ApprovalMax is a paid approval-workflow product with unlimited users and approval steps on its Standard plan, so choose it when approvals are the priority. Doxis targets enterprise content automation with cloud, on-premises or hybrid deployment and custom-fit licensing. Rossum AP Automation is a managed AP platform whose Starter plan begins at $18,000 per year with unlimited seats and API access. Tungsten AP Essentials is a paid AP service with trial access and volume-dependent pricing.

Verdict

invoice2data is the right choice for technically capable teams that want a no-cost, self-hosted invoice extraction component they can shape around their own templates and privacy requirements. Its main reason to choose it is control without licence fees. Look elsewhere when you need managed onboarding, formal support, broad document coverage or a polished end-user workflow rather than an engineering project.

The verdict

invoice2data offers template-based invoice extraction with selectable text and OCR backends, plus optional table and AI features. Its stated scope centers on invoices rather than canonical schemas for other document types.

What the maker publishes

AI options
Optional AI features can draft templates from sample documents or extract fields as a fallback when templates do not match. source
AI providers
The opt-in AI layer lists OpenAI, DeepSeek, Mistral, Gemini, and Ollama providers; Ollama can run locally without sending data off the machine. source
Extraction
It extracts PDF text through selectable backends including pdfium, pdftotext, pdfminer, pdfplumber, and OCR tools. source
Installation
The docs provide pip and conda installation, and state the default pdfium path needs no system libraries on Windows, macOS, or Linux. source
Integration
The project README lists an Odoo OCA module as a user of invoice2data. source
Known scope limit
The README says canonical schemas for non-invoice documents such as waybills, delivery notes, and purchase orders are deferred. source
License
The repository publishes invoice2data under the MIT License, which permits use, modification, distribution, sublicensing, and sale subject to its conditions. source
OCR privacy
The installation guide describes docTR and PaddleOCR as local OCR backends that keep data on the user's machine. source
Outputs
The command-line tool can write extracted results as CSV, JSON, or XML and can rename PDF files to match their contents. source
Privacy behavior
The default AI provider is a mock with no network access, and the docs say no AI runs unless a provider is configured. source
Purpose
invoice2data is a command-line tool and Python library that extracts structured information from PDF invoices for accounting workflows. source
Security
The repository's GitHub security page says no SECURITY.md policy is set up and no security advisories are published. source
Support
The repository invites issues and pull requests, and links to a contributor guide; it does not describe a paid support plan on the pages reviewed. source
Table extraction
Optional plugins support matching invoice line items and tables, including an opt-in Camelot plugin for advanced table parsing. source
Templates
It matches extracted text against YAML or JSON templates that can define custom fields and multiple regexes per field. source

Checked against the maker's pages on 4 Oct 2026.