# Macaron-A2UI: A Model for Generative UI in Personal Agents

arXiv: https://arxiv.org/abs/2605.24830 · HTML: https://mlsanity.com/paper/2605.24830/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 May 2026.

## Built on by (most cited first, 2 of 2)

- [Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA](https://mlsanity.com/paper/2608.09819/) (2026)
- [Task-State Governance: A Virtual Machine for Verifiable and Recoverable GUI Agent Execution](https://mlsanity.com/paper/2610.20100/) (2026)

## Abstract (arXiv)

As personal agents evolve to handle complex, user-centric tasks, static plain-text chat is rapidly becoming a bottleneck. Generative UI emerges as the necessary new interface layer, dynamically synthesizing the right controls, options, and state from the interaction context in real time. We present Macaron-A2UI, a model for Generative UI in personal agents. Our goal is to move beyond text-only interaction by enabling agents to generate natural language together with lightweight, executable UI actions for information collection, preference refinement, confirmation, and multi-goal organization. We build a large-scale Generative UI corpus from heterogeneous dialogue sources, introduce A2UI-Bench for controlled evaluation, and train 30B, 235B and 754B models with parameter-efficient LoRA-based supervised fine-tuning followed by reward-driven reinforcement learning. The best Macaron-A2UI model reaches 75.6 overall on A2UI-Bench without explicit schema hints, surpassing the strongest full-schema frontier baseline. We release the models, benchmark, and evaluation protocol to support future work on Generative UI for personal agents.