Automation
Docling is an open-source document-processing toolkit that converts files into structured content for search and knowledge bases. Its CLI and Python interfaces require installation and the models your pipeline uses.
For developers building knowledge bases or processing documents in batches. Command-line skills and manual checks of difficult scans are needed.
The open-source code is free to use; you provide the runtime and model resources.
Windows
macOS
LinuxPython / CLIPreserve document structure and locate source passages in a retrieval workflow.
Generate descriptions of document pictures to enrich searchable content.
You need Python 3.10+ and command-line skills. Initial runs may download models. A GPU can accelerate supported pipelines, but is not mandatory for every task.
Its main role is parsing and preparing documents. Question answering needs retrieval, storage and a language model; the pictured grounding example uses additional integrations.
Local pipelines can run on your machine. Remote vision models or hosted services may receive the relevant inputs; check your configuration.
These are output examples from official documentation, not Folumi tests. Results vary by model and document.
The open-source code is free to use; you provide the runtime and model resources.
Compute, storage, optional remote models and hosting may cost extra; no hosted allowance is included.
Parse documents with AI to recover layout, tables and picture information.
Visit website ↗Submit once. Get listed permanently after approval.
