In this video I look at SmolDocling and how it compares to the other OCR solutions that are out there, both open and proprietary. 

Blog: https://huggingface.co/blog/smolervlm#smoldocling
Paper:  https://arxiv.org/pdf/2503.11576
HF Model: https://huggingface.co/ds4sd/SmolDocling-256M-preview
Demo: https://huggingface.co/spaces/ds4sd/SmolDocling-256M-Demo

For more tutorials on using LLMs and building agents, check out my Patreon
Patreon: https://www.patreon.com/SamWitteveen
Twitter: https://x.com/Sam_Witteveen

🕵️ Interested in building LLM Agents? Fill out the form below
Building LLM Agents Form: https://drp.li/dIMes

👨‍💻Github:
https://github.com/samwit/llm-tutorials

⏱️Time Stamps:
00:00 SmolDocling Tweet on X
01:35 Docling Github
02:05 SmolDocling Paper
03:26 SmolDocling Hugging Face
03:40 SmolVLM Blog
04:22 SmolDocling Special Instructions
04:29 SmolDocling Demo

Sam Witteveen AI is a publication offering insights, tutorials, and resources for artificial intelligence (AI) enthusiasts and practitioners. Readers can learn about machine learning algorithms, deep learning frameworks, and AI applications. With tutorials, case studies, and expert interviews, Sam Witteveen AI provides  guidance and expertise for building and deploying AI solutions.

Sam Witteveen

SmolDocling is a new OCR model from Hugging Face, developed in collaboration with IBM. It is a compact document understanding model with 256 million parameters, designed to run on GPUs with limited VRAM. It excels in not only OCR but also document conversion, supporting various file types such as PDFs, Microsoft Word files, HTML, and images. Although it may not outperform proprietary models in general OCR tasks, it shows promise when fine-tuned for specific tasks, leveraging its small size and efficiency.

SmolDocling - The SmolOCR Solution?