LogoLogo
WebsiteLoginSign up
  • Welcome to Koncile
  • Getting Started
    • Key Koncile features
    • Folders vs. Extraction Templates
    • General fields vs. Repeated fields
  • DATA EXTRACTION APP
    • Set up an Extraction Template
    • Set field format
    • How to write field descriptions
    • Import documents
      • Importing amendments
    • Export data
    • Auto-classify documents
  • Generated fields with AI
  • API Setup
    • API Key generation
    • Swagger
    • API pricing and limits
    • Libraries
    • File Uploading
    • Task Retrieval
      • Example of Python Code
    • Folders
    • Templates
    • Fields
    • Instructions
    • API status codes
    • Bubble Plugin
  • Miscellaneous
    • Q&A
Powered by GitBook
LogoLogo

Useful links

  • Website
  • Login
  • Extraction template library

@Copyright Koncile 2024

On this page

Was this helpful?

Export as PDF
  1. DATA EXTRACTION APP

Choose the right model

Discover which Koncile models suits your use case

Koncile offers two main extraction models to meet a wide range of use cases.

koncile_universal is a generalist model that leverages computer vision to detect layout and integrates an LLM to interpret meaning with precision. It is the default option available in the Free, Starter, and Advanced plans.

koncile_agentic is a more sophisticated model that incorporates a specialized agent to handle complex information extraction tasks. It is particularly well-suited for:

• Overcoming challenges related to page and line breaks,

• Processing lines with a high density of unit information,

• Managing lengthy documents.

This model ensures superior performance when capturing intricate and structured data.

Last updated 5 months ago

Was this helpful?