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{203821,
        author = {Sakshya Bhattacharya},
        title = {From Text Generators to Autonomous Agents: How Large Language Models Are Being Extended with Tools, Memory, and Planning},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {4926-4935},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=203821},
        abstract = {Not long ago, a large language model was essentially a very sophisticated autocomplete a system that predicted what word should come next, scaled to an almost incomprehensible degree. That description, while technically accurate, no longer captures what these systems actually do. Today's most capable LLMs can browse the web, write and run code, manage files, coordinate with other AI systems, and pursue goals across dozens of sequential steps without a human holding their hand at each turn. This paper traces the technical innovations that made this shift possible. We examine tool-use mechanisms, memory and retrieval architectures, planning and reasoning frameworks, and the emerging ecosystem of multi-agent systems. We take a close look at the platforms that have defined this space LangChain, AutoGPT, Claude Code, and others weighing their design choices honestly. The paper's central argument is simple: the move from passive text generation to goal-directed, tool- equipped agency is not a minor upgrade. It is a fundamental change in what artificial intelligence is for.},
        keywords = {},
        month = {June},
        }