{"slug": "vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation", "title": "VibeVideoBench:Evaluating Multimodal Agents on Code-Driven Video Generation", "summary": "A new benchmark, VibeVideoBench, finds that even the strongest multimodal coding agents fail a core requirement on roughly a quarter of code-driven video generation tasks, according to a paper submitted to arXiv on 26 October 2026. The benchmark spans 239 self-contained tasks in English and Chinese across 12 categories, each requiring an agent in a sandbox to deliver a finished MP4 with sound within one hour, and was scored across nine frontier agent systems and 2,151 rollouts. GPT-6 Astra and Claude Opus 5.5 lead clearly, but three quarters of the videos that were never delivered stemmed from final renders started too late, run too slowly or never started.", "body_md": "# Computer Science > Computer Vision and Pattern Recognition\n\n  [Submitted on 26 Oct 2026]\n\n# Title:VibeVideoBench:Evaluating Multimodal Agents on Code-Driven Video Generation\n\n[View PDF](http://arxiv.org/pdf/2610.35864v1)\n\n[HTML (experimental)](https://arxiv.org/html/2610.35864v1)\n\nAbstract:Diffusion video models produce striking clips, but they cannot reliably render an exact formula, a correct chart, a ruleconsistent game or narration synchronized to on-screen events. A different route is emerging: multimodal agents that write code to render every frame and synthesize every sound. We introduce VibeVideoBench, a benchmark for evaluating agents on code-driven video generation. Its 239 self-contained tasks in English and Chinese span 12 categories, from data stories and music-driven pieces to classroom lessons and paper explainers; each asks an agent, running in its native harness inside a sandbox, to deliver a finished MP4 with sound within one hour. Evaluation is hierarchical and evidence-based: program checks of the delivery spec, about 13k task-specific rubric checks traced to the requirements they enforce, and five quality dimensions anchored per task, judged by three VLM judge panels and validated with blind human preferences. Across nine frontier agent systems and 2,151 scored rollouts, the benchmark is far from solved: GPT-6 Astra and Claude Opus 5.5 lead clearly, yet even the best system fails a core requirement on about a quarter of tasks. Whether a video arrives at all matters as much as its quality, and three quarters of the lost videos are final renders started too late, run too slowly or never started. The judges agree with each other and with human preferences on system rankings, but only moderately on individual videos.\n    \n\n### Additional Features\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation", "canonical_source": "http://arxiv.org/abs/2610.35864v1", "published_at": "2026-10-27 15:31:24+00:00", "updated_at": "2026-10-27 15:43:40.242731+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-research", "generative-ai", "computer-vision"], "entities": ["VibeVideoBench", "arXiv", "GPT-6 Astra", "Claude Opus 5.5"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation", "markdown": "https://wpnews.pro/news/vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation.md", "text": "https://wpnews.pro/news/vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation.txt", "jsonld": "https://wpnews.pro/news/vibevideobench-evaluating-multimodal-agents-on-code-driven-video-generation.jsonld"}}