What is the difference between this and the Microsoft project？https://github.com/microsoft/OmniParser/

[lijwchn]

I am deeply grateful for the author's efforts on this project. I would like to inquire about the difference between this project and the one with Microsoft, purely out of curiosity

https://github.com/microsoft/OmniParser

[adithya-s-k]

Feature/Aspect	OmniParse (Adithya S K)	OmniParser (Microsoft)
Primary Use	General-purpose unstructured data parser for GenAI and LLM applications	User interface (UI) screenshot parser to create actionable structure for computer-using AI agents
Main Input Types	Documents, tables, images, audio, video, web pages	UI screenshots (primarily from computers or mobile devices)
Target Output	Structured, AI/LLM-ready data (e.g., markdown, extracted tables, captions, transcripts)	List of interface elements (buttons, icons) with positions and semantic labels
Primary Audience	Developers building data ingestion pipelines for GenAI, data scientists	AI agents, UI/UX automation, LLM-based computer control
Key Technologies	Python, Docker, supports local and cloud execution	Vision models, fine-tuned YOLOv8, BLIP-2, integration with Florence 2 captioning model