{"paper":{"title":"UIBenchKit: A unified toolkit for design-to-code model evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"UIBenchKit gives researchers a single platform to run and compare design-to-code models under identical conditions.","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Chinh T. Le, Jingyu Xiao, Trevor Ong Yee Siang, Yintong Huo, Yuxuan Wan","submitted_at":"2026-05-13T08:06:41Z","abstract_excerpt":"Recent years have seen substantial progress in automated design-to-code generation, with many methods proposed for generating HTML and CSS from webpage screenshots. However, the absence of a standardized evaluation platform makes it difficult to compare these methods fairly, limiting both practical adoption and systematic research progress. To bridge this gap, we introduce UIBenchKit, an open-source, integrated toolkit designed to unify the evaluation of design-to-code tasks. UIBenchKit abstracts the complexities of environment setup, model inference, and code rendering, offering researchers a"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"UIBenchKit abstracts the complexities of environment setup, model inference, and code rendering, offering researchers a plug-and-play architecture to compare various methods under consistent settings.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the chosen metrics and analytical interface meaningfully capture real-world quality differences between generated code outputs.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"UIBenchKit is a unified open-source toolkit providing plug-and-play evaluation for design-to-code generation models with consistent settings and multi-metric analysis.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"UIBenchKit gives researchers a single platform to run and compare design-to-code models under identical conditions.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f2c38f8c27f5227a46c99565098798848b09c2cdc988c439f7bee91c00d49f45"},"source":{"id":"2605.13141","kind":"arxiv","version":1},"verdict":{"id":"13e9ded0-04ad-4a96-a4cd-59d3a1c6de4b","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T18:28:47.447131Z","strongest_claim":"UIBenchKit abstracts the complexities of environment setup, model inference, and code rendering, offering researchers a plug-and-play architecture to compare various methods under consistent settings.","one_line_summary":"UIBenchKit is a unified open-source toolkit providing plug-and-play evaluation for design-to-code generation models with consistent settings and multi-metric analysis.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the chosen metrics and analytical interface meaningfully capture real-world quality differences between generated code outputs.","pith_extraction_headline":"UIBenchKit gives researchers a single platform to run and compare design-to-code models under identical conditions."},"references":{"count":14,"sample":[{"doi":"","year":2026,"title":"Max Bachmann. 2026. Levenshtein Distance — RapidFuzz Documentation. https: //rapidfuzz.github.io/Levenshtein/levenshtein.html Accessed: 2026-02-26","work_id":"bef538b6-3e6c-4ab5-9ec8-226681baf1aa","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Yi Gui, Zhen Li, Zhongyi Zhang, Yao Wan, Dongping Chen, Hongyu Zhang, Yi Su, Bohua Chen, Xing Zhou, Wenbin Jiang, and Xiangliang Zhang. 2025. UICopilot: Automating UI Synthesis via Hierarchical Code G","work_id":"8996b873-4dd9-43bc-9894-f21b647ecd45","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Yi Gui, Zhen Li, Zhongyi Zhang, Guohao Wang, Tianpeng Lv, Gaoyang Jiang, Yi Liu, Dongping Chen, Yao Wan, Hongyu Zhang, Wenbin Jiang, Xuanhua Shi, and Hai Jin. 2025. <scp>LaTCoder:</scp> Converting Web","work_id":"939f4b56-1ee3-44a2-967e-a103ac64fa69","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Jason Lemkin. 2025. How Vercel Hit $9.3B and Replit Hit $3B. . . After a Decade: The Long Paths to AI Overnight Success. https://www.saastr.com/how-vercel- hit-9-3b-and-replit-hit-3b-after-a-decade-th","work_id":"7710b266-f134-4b43-81c4-8af7878cc48d","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"Meta Open Source. [n. d.]. React Documentation. https://react.dev/. Accessed: 2026-05-10","work_id":"9c35f327-9460-47e4-a364-eb8f3c1d70f5","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":14,"snapshot_sha256":"533df2bedf79c760985f155f57974adfe1110cfea5c367607d70d37d601fd560","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}