arXiv cs.AI· JunJia Guo (Joe), Yuhang Yao (Joe), Chang Li (Joe), Zimeng Pan (Joe), Jiawei (Joe), Zhou, Jingdi Chen·· 9 天前AI 评分41
VISTA：面向视觉设计稿到 Web 应用编码智能体的端到端基准
VISTA: An End-to-End Benchmark for Visual Spec-to-Web-App Coding Agents
AI 导读
VISTA 是面向编码智能体的端到端基准，将多页设计稿（Figma 渲染、结构与文本需求）转化为可运行的全栈 Web 与 Android 应用。它覆盖 18 个应用的 8,487 个人工标注交互组件（126 个 Web 页面、576 个 Android 屏幕），并提供可执行评测器给出定位、行为与联合得分。
正文 · 原文
Abstract:Coding agents can now build applications from design mockups, but a screen that looks right is not an application that works. Existing benchmarks either score visual reconstruction without a backend or test full-stack functionality from textual requirements, and few tie scores to the individual components a design requires. We introduce VISTA (VIsual Spec-To-App), an end-to-end benchmark in which coding agents turn multi-page design handoffs (Figma renders, structure, and textual requirements) into runnable full-stack Web and Mobile (Android) applications. VISTA provides 8,487 human-annotated interactive components across 18 applications (126 Web pages and 576 Android screens) and an executable evaluator that locates annotated components in the running application and probes their interactions, yielding localization, behavior, and joint scores traceable to individual design requirements. Across 14 deployed coding-agent systems, the best joint score is 0.553 on Web and 0.378 on Mobile, and 13 of 14 systems score lower on behavior than on localization on Web. Component probes detect unresponsive controls in 90.1% of audited Web deliveries. Development traces show where the process departs from the task: 44.5% of trajectories never open the design screenshots, 73.0% omit the prescribed self-audit, and 8.3% report success after a failed check. VISTA makes the gap between rendering a design and delivering a working application measurable at the level of individual components. Code is available at \href{this https URL}.
| Comments: | Project page: this https URL |
| Subjects: | Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2605.26144 [cs.SE] |
| (or arXiv:2605.26144v4 [cs.SE] for this version) | |
| https://doi.org/10.48550/arXiv.2605.26144 arXiv-issued DOI via DataCite |
Submission history
From: Junjia Guo [view email]
[v1]
Fri, 22 May 2026 20:29:12 UTC (3,170 KB)
[v2]
Thu, 11 Jun 2026 00:11:54 UTC (3,170 KB)
[v3]
Mon, 22 Jun 2026 18:12:46 UTC (3,172 KB)
[v4]
Tue, 29 Sep 2026 00:12:17 UTC (3,616 KB)
来源：arXiv cs.AI · arxiv.org