[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-bytedance-seedance-2-0-audio-video-unified":3,"topics-all":33,"news-related-d7806d21-255d-4b36-a816-03dfe0fe1030":52},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":20,"news_slug":26,"published_at":27,"created_at":28,"modified_at":29,"is_published":30,"publish_type":31,"image_url":13,"view_count":32},"d7806d21-255d-4b36-a816-03dfe0fe1030","Seedance 2.0：字节跳动统一音视频生成架构的技术突破","2026年2月，字节跳动发布Seedance 2.0，其核心创新在于统一多模态音视频联合生成架构，支持文本、图像、音频、视频四种模态混合输入。模型支持同时输入最多9张图像、3段视频、3段音频片段并配合自然语言控制，这一设计突破了传统视频生成仅支持文本单模态的限制。\n\n音频与视频在统一架构下联合训练而非外部拼接，使生成的视听内容具有内在协调性。官方示例中，双人花样滑冰场景的起跳、旋转、落地等动作序列不仅运动稳定，还严格遵循物理规律，多主体交互时的角色外观一致性也得到保障。\n\n规格层面，Seedance 2.0支持最长15秒多镜头输出并配备双声道音频，为叙事短片创作提供了基础能力。用户可通过自然语言指令精准控制构图、镜头运动与视觉风格，降低了AI视频的使用门槛。\n\n整体来看，这代模型代表了视频生成从单模态向多模态协同演进的重要方向。视觉与听觉信号在模型内部实现深度融合，而非外部模块拼接，是生成内容达到工业可用标准的关键所在。","https:\u002F\u002Farxiv.org\u002Fabs\u002F2604.14148","7d4793fe-c2cb-4759-a1a9-a0e78232593d",[10,14,17],{"id":11,"name":12,"slug":12,"description":13,"color":13},"e676a5cf-1f24-472f-a765-86fa21a1bc3c","ai-model",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"499f4b56-819d-49a3-9609-33e775143b86","multimodal",{"id":18,"name":19,"slug":19,"description":13,"color":13},"ebe5dcd1-46b1-4298-b8c2-8e0e2f456e56","video-generation",[21],{"id":22,"lang":23,"title":24,"summary":25,"content":13},"42736c56-e6d8-46db-83a3-85908f76cfba","en","Seedance 2.0: ByteDance unifies audio and video generation","In February 2026, ByteDance released Seedance 2.0, whose core innovation is the unified multimodal audio-video joint generation architecture, supporting text, image, audio, and video as four mixed-input modalities. The model supports simultaneously inputting up to 9 images, 3 video clips, 3 audio segments combined with natural-language control — a design that breaks the limitation of traditional video generation only supporting text single-modality.\n\nAudio and video are jointly trained within a unified architecture rather than externally stitched, giving the generated audiovisual content inherent coordination. In the official example, the figure-skating scene's take-off, rotation, and landing motion sequence is not only motionally stable but also strictly follows physical laws, with character appearance consistency preserved in multi-subject interactions.\n\nOn specs, Seedance 2.0 supports up to 15-second multi-shot output with dual-channel audio, providing foundational capability for narrative short-film creation. Users can precisely control composition, camera movement, and visual style through natural-language instructions, lowering the bar for AI video use.\n\nOverall, this generation of model represents an important direction in video generation's evolution from single-modality to multimodal collaboration. The deep fusion of visual and auditory signals inside the model — rather than external module stitching — is key to generated content reaching industrial-usable standards.","bytedance-seedance-2-0-audio-video-unified","2026-05-04T05:01:00Z","2026-05-04T13:09:25.808344Z","2026-08-19T02:08:40.142862Z",true,"agent",192,[34,43],{"slug":35,"tag_slug":35,"title_zh":36,"title_en":37,"intro_zh":38,"intro_en":39,"id":40,"is_active":30,"created_at":41,"modified_at":42},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":44,"tag_slug":44,"title_zh":45,"title_en":46,"intro_zh":47,"intro_en":48,"id":49,"is_active":30,"created_at":50,"modified_at":51},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":53},[54,59,64,69,74,79],{"id":55,"title":56,"news_slug":57,"published_at":58},"25eeddf5-b3d7-45b4-9725-387f60955252","智象未来提出多模态模型Token利润率远超语言模型：世界模型的技术路径与商业化前景","zhixiang-multimodal-token-margin-world-model","2026-05-28T10:00:00+00:00",{"id":60,"title":61,"news_slug":62,"published_at":63},"ff7fa3e1-6737-4bd9-85c8-8010d13a44f3","字节跳动 Bernini 开源：用 MLLM 当\"语义规划师\"，拆开视频生成的\"思考\"与\"渲染\"","bytedance-bernini-mllm-semantic-planner-dit","2026-05-21T00:00:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"b7e375b3-6949-4ff6-9cd4-012aced6bbc3","Runway赌注视频生成通往世界模型：挑战Google的架构之赌","runway-video-world-model-google-bet","2026-05-16T04:10:00+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"8c908abd-6e59-4ccc-949d-0876c9dfbc2d","Runway Gen-4 发布：原生音视频同步 + 物理引擎升级，视频生成进入新阶段","runway-gen-4-audio-video-sync-physics","2026-05-08T16:10:00+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"dc534f1a-dbe2-4f79-9557-bbb43a82f24f","视频模型总在绕过 MLLM,三段训练把它拉回来","fapl-mllm-latents-cinematic-video-generation","2026-10-23T09:20:00+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"cf7aeaaa-7179-4de2-b177-9adf54668d61","StreamEval：37 模型实测，实时视频最缺听指令","streameval-interactive-streaming-video-benchmark","2026-10-22T11:08:09+00:00"]