# Open WebUI: An Open, Extensible, and Usable Interface for AI Interaction

arXiv: https://arxiv.org/abs/2510.02546 · HTML: https://mlsanity.com/paper/2510.02546/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Oct 2025.

## Built on by (most cited first, 6 of 6)

- [Secure On-Premise Deployment of Open-Weights Large Language Models in Radiology: An Isolation-First Architecture with Prospective Pilot Evaluation](https://mlsanity.com/paper/2604.22768/) (2026)
- [VectorizationLLM: Smart Vectorization Based AI Assistant](https://mlsanity.com/paper/2607.07846/) (2026)
- [A corrective agentic hybrid RAG and an operations-grounded evaluation for a scientific facility](https://mlsanity.com/paper/2607.24663/) (2026)
- [Carefully Considering Culture: Analyzing LLM Alignment in Single- and Multi-Cultural Settings using Cultural Consensus Theory](https://mlsanity.com/paper/2608.09937/) (2026)
- [Physico-Geospatial Grounded Scene Interpretation for Mobile Robotics](https://mlsanity.com/paper/2609.06629/) (2026)
- [CultureBot: A Framework for Retrieval-Augmented Exploration of Cultural Heritage Data](https://mlsanity.com/paper/2610.16435/) (2026)

## Abstract (arXiv)

While LLMs enable a range of AI applications, interacting with multiple models and customizing workflows can be challenging, and existing LLM interfaces offer limited support for collaborative extension or real-world evaluation. In this work, we present an interface toolkit for LLMs designed to be open (open-source and local), extensible (plugin support and users can interact with multiple models), and usable. The extensibility is enabled through a two-pronged plugin architecture and a community platform for sharing, importing, and adapting extensions. To evaluate the system, we analyzed organic engagement through social platforms, conducted a user survey, and provided notable examples of the toolkit in the wild. Through studying how users engage with and extend the toolkit, we show how extensible, open LLM interfaces provide both functional and social value, and highlight opportunities for future HCI work on designing LLM toolkit platforms and shaping local LLM-user interaction.