Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM61935
Volume:
10 |
4 |
April 2026
Pages:
1-3
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
Department of Computer Science
Rizvi College of Engineering Mumbai, India
Department of Computer Science
Rizvi College of Engineering Mumbai, India
Department of Computer Science
Rizvi College of Engineering Mumbai, India
This paper presents Vibe, an AI-assisted web development system designed to automate the process of creating fully functional web applications through natural language conversation. Traditional web development requires extensive expertise in HTML, CSS, JavaScript, and modern frameworks, making it inaccessible to non-technical users. Existing no-code and low-code platforms, while simplifying the process, impose rigid templates and limited customization, failing to deliver production-quality, framework-native output. The proposed system integrates a multi-agent AI architecture, sandboxed code execution, and real-time preview capabilities to enable end-to-end website generation. It accepts natural language prompts via a conversational interface, processes them through specialized AI agents built on the Inngest Agent Kit with OpenAI's GPT-4.1, and generates complete Next.js applications within isolated E2B cloud sandboxes. Experimental evaluation demonstrates that the system successfully generates complex, multi-component web applications with realistic interactivity and production-quality code structure, reducing development time from hours to minutes.
KEYWORDS
Vibe demonstrates how AI-assisted systems can significantly improve web development workflows by automating the entire process from natural language description to functional web application. The system successfully integrates multi-agent AI architecture, sandboxed code execution, and real-time preview capabilities to produce production-quality Next.js applications. The multi-agent approach proves effective in decomposing complex generation tasks while maintaining output quality. The system reduces development time from hours or days to minutes, democratizing access to modern web application creation for users regardless of their technical background.
[1] M. Chen et al., "Evaluating Large Language Models Trained on Code," arXiv preprint arXiv:2107.03374, 2021. [2] OpenAI, "GPT-4 Technical Report," arXiv preprint arXiv:2303.08774, 2023. [3] R. Li et al., "StarCoder: May the Source Be with You!" arXiv preprint arXiv:2305.06161, 2023. [4] S. Hong et al., "MetaGPT: Meta Programming for Multi-Agent Collaborative Framework," arXiv preprint arXiv:2308.00352, 2023.
Publication Date
5/1/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
2
Total views since publication
PDF Downloads
4
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...