Skip to content

granite docling 258M

ibm-granite/granite-docling-258M
Open Source · chat · open-weights
GA Alert me on changes
Context
8K
Max output
Pricing
Open weights
Modalities
text · image
Released
19 May 2025
License: apache-2.0 · ibm-granite/granite-docling-258M
AI summary
● machine-written

IBM Granite Docling 258M: Open-source document AI model

Granite Docling 258M is a 258-million parameter vision-language model designed for document conversion and structured extraction. It integrates vision and language capabilities to parse PDFs, slides, and scanned pages into machine-readable formats while preserving layout, tables, code, and equations. The model is released open-source under Apache 2.0 and succeeds SmolDocling with improvements in OCR, equation handling, and table accuracy.

What's new
  • 258M parameter compact vision-language model for document conversion
  • Integrates vision and language in single model instead of chained OCR pipeline
  • Outputs structured DocTags to preserve layout, tables, code, and equations
  • Supports English natively with experimental Japanese, Arabic, and Chinese support
Best for
Document conversion and extraction workflows (PDF to Markdown/HTML)Parsing tables, code blocks, and mathematical equations from documentsLayout-preserving document digitization and structured data extraction
Sources

Source: https://huggingface.co/ibm-granite/granite-docling-258M