What you need to know about Lector
Pricing
Experience the next generation of document intelligence at no cost!
0 EUR
100 free AI page processings (standard models)
Available once only
Access to all available AI features
Perfect for individuals and teams optimising workflows without any commitment
Pay-per-use
Pay as you go with monthly billing
4 cents per AI page processing (standard models)
Cancel any time
Response time for incidents: 5 working days
2 bookable consulting hours per month
Scale document automation for companies of any size
Business S Monthly: 599.00 EURBusiness S Yearly: 499.00 EURBusiness M Monthly: 1,199.00 EURBusiness M Yearly: 999.00 EURBusiness L Monthly: 2,399.00 EURBusiness L Yearly: 1,999.00 EUR
per monthper month · billed yearly
Business S: 16,000 free AI page processings per month (standard models)Business M: 42,000 free AI page processings per month (standard models)Business L: 125,000 free AI page processings per month (standard models)
Business S: 3.2 cents per additional AI page processing (standard models)Business M: 2.4 cents per additional AI page processing (standard models)Business L: 1.6 cents per additional AI page processing (standard models)
Business S: Response time for incidents: 3 working daysBusiness M: Response time for incidents: 2 working daysBusiness L: Response time for incidents: 8 hours
Business S: 4 bookable consulting hours per monthBusiness M: 8 bookable consulting hours per monthBusiness L: 16 bookable consulting hours per month
Optional: premium SLA
Tailor-made solutions for large organisations
Custom
Custom page volumes
Custom contract terms
Premium SLA included
Custom response times
Unlimited bookable consulting hours
Custom AI models
Custom import and export channels
Integration with external identity providers
Quick start
Set up complex workflows in next to no time
Powered by state-of-the-art generative AI models
No separator sheets or stamps required
Reliable recognition even for photos & handwriting
New document class
Ready to process
Invoice Fields & descriptions
Precise processing of complex, unstructured documents
Recognition of handwritten content, too
Multimodal models that process both text and images
Quickly verify the results on your own documents
Refine your description in short iteration cycles
Evaluate accuracy against your own data
Inbox · Invoice
3 fields extracted
Enterprise
Lector is built from the ground up for enterprise use — with everything large organisations need.
Multi-tenancy
Granular role and permission management
SLA agreements
Dedicated onboarding and support
Custom model adaptation
Millions of documents per month
Monitoring & insights
Consulting & integration
Not a self-service project that ends up on your desk: our team sets up document classes, fields and the connection to your systems together with you.
Kick-off, joint configuration and a test run with your real documents — before the first batch goes live. All done by our own team, no offshoring.
Tell us your systems, document types and volumes. For integration projects you get a concrete quote at short notice — within three working days at the latest.
Case-management systems, DMS/ECM, ERP, mailboxes, RPA: we have connected Lector to many system landscapes — from public authorities and health insurers to banks. Afterwards, the consulting hours in your plan keep you covered.
Integrations
No system migration. No parallel operation. Lector integrates with your existing infrastructure.
Input
Classify · ExtractOutput
Frequently asked
Invoices, forms, letters, emails, images and tables — as PDF, DOC, DOCX, RTF, ODT, XLS, XLSX, ODS, CSV, JPEG, PNG, TIFF, HEIC, HEIF, HTML, XML, Markdown, MSG, EML, vCard, ZIP or TXT.
Currently Gemini, GPT, Claude, Mistral, InternVL and Qwen — open and closed source. Closed-source models run GDPR-compliant exclusively on European servers (Google Vertex AI, Microsoft Azure). For enterprise customers we additionally develop custom models.
Import via REST API, email, SFTP/file system, messaging/webhook or from your own apps. Export via webhook or API by default, plus DMS/ECM/archive, ERP, databases, cloud storage, email and RPA tools (UiPath, Power Automate). All details are in the documentation.
A system running Ubuntu with at least 16 CPU cores (32+ recommended), at least 32 GB RAM (64 GB recommended) and around 1 TB of storage. AI processing requires a GPU, e.g. Nvidia H100 (80 GB) or RTX 6000 Ada (48 GB).
From just a few documents a month: the Starter plan bills per processed page with no base fee. From roughly 15,000 AI page calls a month, Business S is cheaper than pay-per-use.
Yes. The free trial includes 100 AI page processings with access to all AI features — once, with no commitment.
Every processing step is traceable: Lector keeps an immutable history per document (audit trail), shows straight-through-processing rates per document class and provides usage statistics for system and AI components. Confidence values are provided for certain AI models. KPIs can be pulled into your own reporting via GraphQL, system metrics monitored via Prometheus. For quality itself there is the Evaluation Center: compare settings and models on real documents before they go live.
