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{186681,
        author = {Gorhe Varad Baban and Jadhav Rohit Balasaheb and Kasar Amit Sudhakar and Prof P S Bramhane},
        title = {AI POWERED API TESTING},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {6},
        pages = {2262-2267},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=186681},
        abstract = {Application Programming Interfaces (APIs) are fundamental to modern distributed software architectures, yet traditional testing approaches struggle with inefficiency, limited coverage, and high maintenance costs . This paper presents a novel AI-powered web application framework that automates RESTful API testing by integrating Retrieval-Augmented Generation (RAG) with locally deployed Large Language Models (LLMs) to dynamically generate, execute, and validate JSON test payloads . The system addresses critical limitations including manual effort, brittle scripts, and incomplete edge case detection by storing OpenAPI specifications in a vector-based knowledge base and using context-aware generation to produce diverse test scenarios covering positive, negative, and boundary conditions . Implementation leverages FastAPI backend with React.js frontend, supporting parallel execution with dependency resolution, JSON diff validation, and Broken Object Level Authorization (BOLA) security checks via Karate DSL integration . Experimental evaluation using mock APIs demonstrates 85% path coverage, 95% code-less automation, and 60-80% reduction in manual testing cycles within sub-150-second latency on standard 8GB RAM hardware . The framework employs SQLite for relational data persistence and ChromaDB for RAG embeddings, ensuring offline operation with 99% uptime and AES-256 data encryption . Results validate the approach's effectiveness in enhancing defect detection by 30-50% through intelligent scenario exploration while maintaining WCAG 2.1 AA accessibility standards . This work establishes a scalable, privacy-preserving model for AI-driven quality assurance, with future extensions targeting GraphQL support, cloud-native deployment, and self-healing test capabilities .},
        keywords = {API testing, Retrieval-Augmented Generation, Large Language Model, Automation, Security Testing, Code-less Testing},
        month = {November},
        }