Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{199761,
        author = {Sravani Bondalapati and Dr. G. Ramaswamy},
        title = {OCR-Free Desktop AI Assistant for Real-Time Writing and Coding Support Using Dynamically Selected LLMs},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {13834-13849},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=199761},
        abstract = {Typically, desktop users may employ several apps, including grammar corrector tools, content improvers, code debuggers, and coding aids, which require switching from one app to another, causing interruption of workflow, increased efforts, and decreased efficiency. Most current desktop assistive technologies employ screen shotting and OCR to capture screen content, which decreases their efficiency and accuracy of operation. This paper introduces the Desktop AI Assistant, an application providing immediate and contextualized suggestions inside active desktop applications without requiring users to change applications. Instead of OCR, it utilizes direct file handling techniques to access the active file, increasing the efficiency and accuracy of suggestion generation. To recognize the nature of the document (text vs code), a rule-based classifier is used. Accordingly, the Desktop AI Assistant selects the appropriate Large Language Model (LLM) to offer context-sensitive suggestions. DeepSeek R1 is used to detect grammar mistakes and improve content quality, while Code Llama can detect bugs, check the syntax, and optimize the code. The system is implemented using Python with PyQt or Electron, enabling integration with applications such as Microsoft Word, VS Code, PyCharm, and Notepad. Experimental results show improved efficiency and accuracy compared to OCR-based systems.},
        keywords = {Desktop AI Assistant, Large Language Models (LLMs), Context-Aware Computing, Code Debugging, Grammar Correction, Real-Time Suggestions},
        month = {April},
        }