{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-07-28T05:20:57.982Z","headline":"Black Forest Labs 发布 FLUX 3 多模态模型，支持单次生成 20 秒视频与原生音频","description":"Black Forest Labs 以 Early Access 方式推出 FLUX 3 多模态基础模型，采用统一架构联合学习图像、视频和音频。该模型基于 Self-Flow 学习框架扩展，可在单次生成中输出最长 20 秒视频并附带原生音频，支持文生视频、图生视频、多镜头串联等任务。","url":"https://www.aioga.com/news/cmrykywyx026jrolgq06set4j/","mainEntityOfPage":"https://www.aioga.com/news/cmrykywyx026jrolgq06set4j/","datePublished":"2026-07-24T06:48:04.000Z","dateModified":"2026-07-24T06:48:04.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://www.ithome.com/0/981/137.htm","https://aihot.virxact.com/items/cmrykywyx026jrolgq06set4j"],"canonicalUrl":"https://www.aioga.com/news/cmrykywyx026jrolgq06set4j/","directAnswer":{"@type":"Answer","text":"Black Forest Labs 以 Early Access 方式推出 FLUX 3。该模型通过统一架构联合学习图像、视频和音频，单次可生成最长 20 秒并带原生音频的视频，支持文生视频、图生视频、多镜头串联等任务。","url":"https://www.aioga.com/news/cmrykywyx026jrolgq06set4j/","dateCreated":"2026-07-24T06:48:04.000Z","author":{"@type":"Organization","@id":"https://www.aioga.com/authors/aioga-editorial/#editorial-team","name":"Aioga Editorial Team","url":"https://www.aioga.com/authors/aioga-editorial/"}},"evidence":[{"@type":"CreativeWork","name":"IT之家 source article","url":"https://www.ithome.com/0/981/137.htm","datePublished":"2026-07-24T06:48:04.000Z","provider":{"@type":"Organization","name":"IT之家","url":"https://www.ithome.com/0/981/137.htm"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cmrykywyx026jrolgq06set4j","datePublished":"2026-07-24T06:48:04.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cmrykywyx026jrolgq06set4j"}}],"aggregationSource":"IT之家（RSS）","originalPublisher":{"name":"IT之家","url":"https://www.ithome.com/0/981/137.htm"},"article":{"id":"cmrykywyx026jrolgq06set4j","slug":"cmrykywyx026jrolgq06set4j","url":"https://www.aioga.com/news/cmrykywyx026jrolgq06set4j/","title":"Black Forest Labs 发布 FLUX 3 多模态模型，支持单次生成 20 秒视频与原生音频","title_en":"FLUX 3 多模态 AI 模型登场：支持单次生成 20 秒多样化视频","summary":"Black Forest Labs 以 Early Access 方式推出 FLUX 3 多模态基础模型，采用统一架构联合学习图像、视频和音频。该模型基于 Self-Flow 学习框架扩展，可在单次生成中输出最长 20 秒视频并附带原生音频，支持文生视频、图生视频、多镜头串联等任务。","source":"IT之家（RSS）","sourceUrl":"https://www.ithome.com/0/981/137.htm","aiHotUrl":"https://aihot.virxact.com/items/cmrykywyx026jrolgq06set4j","publishedAt":"2026-07-24T06:48:04.000Z","category":"模型更新","score":73,"selected":true,"articleBody":["IT之家：https://www.ithome.com/ 7 月 24 日消息，Black Forest Labs 昨日（7 月 23 日）发布博文，宣布以 Early Access 方式，上线推出 Flux 3 多模态基础模型， 采用统一架构联合学习图像、视频和音频。","在训练方面，IT之家援引博文介绍，该模型基于 Self-Flow（自监督流匹配）学习框架扩展，同步训练视频、图像和音频，在 FLUX.1 和 FLUX.2 系列基础上，扩大至多模态生成与理解任务。","FLUX 3 可在单次生成中输出最长 20 秒的视频，并附带原生音频。官方表示该模型支持文生视频、图生视频、视频生视频、输入视频与音频续写、关键帧转视频、多语言对话，以及多镜头串联。","性能方面，在带声音的 10 秒、720p 视频人工评测方面，结果显示，FLUX 3 对比 Grok Imagine Video 的胜率为 69%，对比 Seedance 2.0 的胜率为 52%，对比 Gemini Omni Flash 的胜率也为 52%。","在机器人方向，Black Forest Labs 正与 Mimic Robotics 合作，研究将 FLUX 3 用作机器人行为预测模型。","图像能力方面，官方称 FLUX 3 可完成图像生成与编辑，覆盖多种风格、宽高比和分辨率。"],"articleImages":[{"sourceUrl":"https://img.ithome.com/newsuploadfiles/2026/7/dcf06bb9-348d-4bc3-8577-8f4122e87e5b.png","alt":"","afterParagraph":1,"url":"/media/articles/cmrykywyx026jrolgq06set4j/2517d17acb79d716.png"},{"sourceUrl":"https://img.ithome.com/newsuploadfiles/2026/7/34a64589-87ab-4f58-80af-7034c41f58cb.png","alt":"","afterParagraph":1,"url":"/media/articles/cmrykywyx026jrolgq06set4j/bd3d791403ffad6a.png"},{"sourceUrl":"https://img.ithome.com/newsuploadfiles/2026/7/cfceb9c8-ad33-49eb-9ada-98077ed24e9b.png","alt":"","afterParagraph":2,"url":"/media/articles/cmrykywyx026jrolgq06set4j/5d34d10b6d2a2759.png"},{"sourceUrl":"https://img.ithome.com/newsuploadfiles/2026/7/21032ba6-71b6-45cd-aa3f-1e0e63992dc9.png","alt":"","afterParagraph":3,"url":"/media/articles/cmrykywyx026jrolgq06set4j/b05aaf3e4a3a08fc.png"},{"sourceUrl":"https://img.ithome.com/newsuploadfiles/2026/7/a374a041-83a5-4337-9dda-302f43c650ce.png","alt":"","afterParagraph":5,"url":"/media/articles/cmrykywyx026jrolgq06set4j/90731a482370411c.png"}],"mediaStatus":"ok","articleBodyZh":[],"translationStatus":"","bodyOrigin":"source-page","editorial":{"summary":"Black Forest Labs 以 Early Access 方式推出 FLUX 3。该模型通过统一架构联合学习图像、视频和音频，单次可生成最长 20 秒并带原生音频的视频，支持文生视频、图生视频、多镜头串联等任务。","background":"公开材料显示，FLUX 3 基于 Self-Flow 学习框架扩展，在 FLUX.1 和 FLUX.2 系列基础上覆盖多模态生成与理解。官方还列出视频生视频、音视频续写、关键帧转视频和多语言对话等能力。","viewpoint":"Aioga 判断，统一处理图像、视频与音频是此次更新的核心看点，20 秒单次生成和原生音频有助于减少分段制作环节。不过模型仍处于 Early Access 阶段，其稳定性与实际工作流表现仍值得关注。","implications":"该模型可能推动视频生成从单一画面输出转向音画联合生成，并为多镜头内容制作提供新的实现方式。官方人工评测给出了与三款模型的对比胜率，但这些结果仍需结合评测条件和独立测试理解。","nextStep":"后续应关注 Early Access 的开放范围、实际生成质量与任务稳定性，并核验其在不同语言、分辨率、多镜头衔接和音视频续写场景中的表现。机器人行为预测合作的研究进展也值得持续跟踪。","evidenceRefs":["title","summary","articleBody","source"],"status":"published","aiGenerated":true,"autoApproved":true,"generatedBy":"aioga-editorial:gpt-5.6-sol","reviewedBy":"aioga-editorial-review:gpt-5.6-sol","generatedAt":"2026-07-26T07:49:35.502Z","sourceHash":"febc1b6ba6ba0d81","review":{"approved":true,"groundedness":94,"clarity":91,"duplicationRisk":24,"blockingIssues":[],"notes":["“20 秒单次生成和原生音频有助于减少分段制作环节”属于合理推断，已通过“Aioga 判断”明确标注为观点。","“可能推动视频生成从单一画面输出转向音画联合生成”属于趋势判断，使用了不确定性表述，没有冒充既成事实。","候选内容对官方人工评测保持了必要限定，未直接将官方胜率结论扩大为普遍性能优势。"]},"validation":{"passed":true,"mode":"ai-auto","revisions":0,"checks":["schema","length","source-attribution","low-source-overlap","no-html","independent-ai-review"]}},"tags":["模型更新","IT之家（RSS）"],"translations":{"zh-CN":{"title":"Black Forest Labs 发布 FLUX 3 多模态模型，支持单次生成 20 秒视频与原生音频","summary":"Black Forest Labs 以 Early Access 方式推出 FLUX 3 多模态基础模型，采用统一架构联合学习图像、视频和音频。该模型基于 Self-Flow 学习框架扩展，可在单次生成中输出最长 20 秒视频并附带原生音频，支持文生视频、图生视频、多镜头串联等任务。","category":"模型更新","source":"IT之家","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs 发布 FLUX 3 多模态模型，支持单次生成 20 秒视频与原生音频 - Aioga AI资讯","description":"Black Forest Labs 以 Early Access 方式推出 FLUX 3 多模态基础模型，采用统一架构联合学习图像、视频和音频。该模型基于 Self-Flow 学习框架扩展，可在单次生成中输出最长 20 秒视频并附带原生音频，支持文生视频、图生视频、多镜头串联等任务。","url":"https://www.aioga.com/news/cmrykywyx026jrolgq06set4j/"},"en":{"title":"Black Forest Labs releases the FLUX 3 multimodal model, supporting single-generation of 20-second videos with native audio","summary":"Black Forest Labs has launched the FLUX 3 multimodal foundation model in Early Access, using a unified architecture to jointly learn images, videos, and audio. This model is extended based on the Self-Flow learning framework and can output videos up to 20 seconds long with native audio in a single generation, supporting tasks such as text-to-video, image-to-video, and multi-shot concatenation.","category":"Models","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs releases the FLUX 3 multimodal model, supporting single-generation of 20-second videos with native audio - Aioga AI News","description":"Black Forest Labs has launched the FLUX 3 multimodal foundation model in Early Access, using a unified architecture to jointly learn images, videos, and audio. This model is extend...","url":"https://www.aioga.com/en/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:25:11.599Z"},"ja":{"title":"Black Forest Labs、FLUX 3 マルチモーダルモデルを発表、20秒の動画とネイティブオーディオの単回生成をサポート","summary":"Black Forest Labs は Early Access 方式で FLUX 3 マルチモーダル基盤モデルをリリースし、統一アーキテクチャで画像、動画、音声を共同学習します。このモデルは Self-Flow 学習フレームワークに基づいて拡張されており、1 回の生成で最長 20 秒の動画を生成し、ネイティブ音声を付加することができます。また、テキストから動画、画像から動画、マルチカット連結などのタスクをサポートします。","category":"モデル更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs、FLUX 3 マルチモーダルモデルを発表、20秒の動画とネイティブオーディオの単回生成をサポート - Aioga AIニュース","description":"Black Forest Labs は Early Access 方式で FLUX 3 マルチモーダル基盤モデルをリリースし、統一アーキテクチャで画像、動画、音声を共同学習します。このモデルは Self-Flow 学習フレームワークに基づいて拡張されており、1 回の生成で最長 20 秒の動画を生成し、ネイティブ音声を付加することができます。また、テキストから...","url":"https://www.aioga.com/ja/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:25:21.980Z"},"ko":{"title":"Black Forest Labs가 FLUX 3 다중 모달 모델을 출시하여 단일 생성으로 20초 비디오와 원본 오디오를 지원합니다","summary":"Black Forest Labs는 Early Access 방식으로 FLUX 3 멀티모달 기본 모델을 출시했으며, 통합 아키텍처를 사용해 이미지, 비디오, 오디오를 공동 학습한다. 이 모델은 Self-Flow 학습 프레임워크를 기반으로 확장되어 한 번의 생성으로 최대 20초 영상과 원본 오디오를 출력할 수 있으며, 텍스트 기반 영상 생성, 이미지 기반 영상 생성, 다중 카메라 연결 등 작업을 지원한다.","category":"모델 업데이트","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs가 FLUX 3 다중 모달 모델을 출시하여 단일 생성으로 20초 비디오와 원본 오디오를 지원합니다 - Aioga AI 뉴스","description":"Black Forest Labs는 Early Access 방식으로 FLUX 3 멀티모달 기본 모델을 출시했으며, 통합 아키텍처를 사용해 이미지, 비디오, 오디오를 공동 학습한다. 이 모델은 Self-Flow 학습 프레임워크를 기반으로 확장되어 한 번의 생성으로 최대 20초 영상과 원본 오디오를 출력할 수 있으며, 텍스트...","url":"https://www.aioga.com/ko/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:25:58.673Z"},"es":{"title":"Black Forest Labs lanza el modelo multimodal FLUX 3, que admite la generación de videos de 20 segundos y audio nativo en una sola vez","summary":"Black Forest Labs lanzó el modelo base multimodal FLUX 3 en acceso anticipado, utilizando una arquitectura unificada para aprender conjuntamente imágenes, videos y audio. Este modelo se basa en la ampliación del marco de aprendizaje Self-Flow, puede generar videos de hasta 20 segundos en una sola vez con audio nativo, y soporta tareas como generación de videos a partir de texto, generación de videos a partir de imágenes y encadenamiento de múltiples tomas.","category":"Modelos","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs lanza el modelo multimodal FLUX 3, que admite la generación de videos de 20 segundos y audio nativo en una sola vez - Aioga Noticias de IA","description":"Black Forest Labs lanzó el modelo base multimodal FLUX 3 en acceso anticipado, utilizando una arquitectura unificada para aprender conjuntamente imágenes, videos y audio. Este mode...","url":"https://www.aioga.com/es/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:25:57.475Z"},"fr":{"title":"Black Forest Labs publie le modèle multimodal FLUX 3, capable de générer en une seule fois une vidéo de 20 secondes avec audio natif","summary":"Black Forest Labs a lancé le modèle fondamental multimodal FLUX 3 en accès anticipé, utilisant une architecture unifiée pour l'apprentissage conjoint d'images, de vidéos et d'audio. Ce modèle est basé sur le cadre d'apprentissage Self-Flow étendu, pouvant produire jusqu'à 20 secondes de vidéo avec audio natif en une seule génération, et prend en charge des tâches telles que la génération de vidéo à partir de texte, la génération de vidéo à partir d'images et la concaténation de plusieurs plans.","category":"Modèles","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs publie le modèle multimodal FLUX 3, capable de générer en une seule fois une vidéo de 20 secondes avec audio natif - Aioga Actualités IA","description":"Black Forest Labs a lancé le modèle fondamental multimodal FLUX 3 en accès anticipé, utilisant une architecture unifiée pour l'apprentissage conjoint d'images, de vidéos et d'audio...","url":"https://www.aioga.com/fr/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:26:34.246Z"},"de":{"title":"Black Forest Labs veröffentlicht das multimodale Modell FLUX 3, das die einmalige Erstellung von 20-Sekunden-Videos mit nativer Audioausgabe unterstützt","summary":"Black Forest Labs hat das multimodale Basis-Modell FLUX 3 im Early-Access-Verfahren eingeführt, das mithilfe einer einheitlichen Architektur Bilder, Videos und Audio gemeinsam lernt. Das Modell basiert auf dem erweiterten Self-Flow-Lernrahmenwerk und kann in einem einzigen Durchlauf Videos von bis zu 20 Sekunden Länge mit nativer Audioausgabe erzeugen. Es unterstützt Aufgaben wie textbasierte Videoerstellung, bildbasierte Videoerstellung und die Verkettung mehrerer Kameraperspektiven.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs veröffentlicht das multimodale Modell FLUX 3, das die einmalige Erstellung von 20-Sekunden-Videos mit nativer Audioausgabe unterstützt - Aioga KI-News","description":"Black Forest Labs hat das multimodale Basis-Modell FLUX 3 im Early-Access-Verfahren eingeführt, das mithilfe einer einheitlichen Architektur Bilder, Videos und Audio gemeinsam lern...","url":"https://www.aioga.com/de/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:26:34.504Z"},"pt-BR":{"title":"A Black Forest Labs lançou o modelo multimodal FLUX 3, que suporta a geração de vídeos de 20 segundos e áudio nativo em uma única execução","summary":"Black Forest Labs lançou o modelo base multimodal FLUX 3 em Acesso Antecipado, usando uma arquitetura unificada para aprender imagens, vídeos e áudios de forma conjunta. Este modelo, baseado na estrutura de aprendizado Self-Flow, é expandido para gerar vídeos de até 20 segundos em uma única vez com áudio nativo, suportando tarefas como vídeos gerados a partir de texto, vídeos a partir de imagens e concatenamento de múltiplas câmeras.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"A Black Forest Labs lançou o modelo multimodal FLUX 3, que suporta a geração de vídeos de 20 segundos e áudio nativo em uma única execução - Aioga Notícias de IA","description":"Black Forest Labs lançou o modelo base multimodal FLUX 3 em Acesso Antecipado, usando uma arquitetura unificada para aprender imagens, vídeos e áudios de forma conjunta. Este model...","url":"https://www.aioga.com/pt-BR/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:27:09.612Z"},"ru":{"title":"Black Forest Labs выпустила многомодальную модель FLUX 3, поддерживающую однократное создание 20-секундного видео с оригинальным звуком","summary":"Black Forest Labs выпустила мульти-модальную базовую модель FLUX 3 в формате раннего доступа, используя унифицированную архитектуру для совместного обучения изображений, видео и аудио. Эта модель основана на расширенном фреймворке Self-Flow, способна выводить видео длиной до 20 секунд с нативным аудио за один раз и поддерживает задачи генерации видео по тексту, генерации видео по изображению, последовательного соединения нескольких сцен и другие задачи.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs выпустила многомодальную модель FLUX 3, поддерживающую однократное создание 20-секундного видео с оригинальным звуком - Aioga Новости ИИ","description":"Black Forest Labs выпустила мульти-модальную базовую модель FLUX 3 в формате раннего доступа, используя унифицированную архитектуру для совместного обучения изображений, видео и ау...","url":"https://www.aioga.com/ru/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:27:17.454Z"},"ar":{"title":"أصدرت Black Forest Labs نموذج FLUX 3 متعدد الوسائط، يدعم إنشاء فيديو مدته 20 ثانية وصوت أصلي في مرة واحدة","summary":"أطلقت Black Forest Labs نموذج FLUX 3 متعدد الوسائط بنظام الوصول المبكر، باستخدام بنية موحدة لتعلم الصور والفيديو والصوت بشكل مشترك. يعتمد هذا النموذج على إطار التعلم Self-Flow الموسع، ويمكنه إنتاج فيديو يصل طوله إلى 20 ثانية مع الصوت الأصلي في مرة واحدة، ويدعم مهام إنشاء الفيديو من النص، إنشاء الفيديو من الصور، وربط لقطات متعددة.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"أصدرت Black Forest Labs نموذج FLUX 3 متعدد الوسائط، يدعم إنشاء فيديو مدته 20 ثانية وصوت أصلي في مرة واحدة - Aioga أخبار الذكاء الاصطناعي","description":"أطلقت Black Forest Labs نموذج FLUX 3 متعدد الوسائط بنظام الوصول المبكر، باستخدام بنية موحدة لتعلم الصور والفيديو والصوت بشكل مشترك. يعتمد هذا النموذج على إطار التعلم Self-Flow المو...","url":"https://www.aioga.com/ar/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:27:58.536Z"},"hi":{"title":"ब्लैक फॉरेस्ट लैब्स ने FLUX 3 मल्टीमॉडल मॉडल जारी किया, जो एक बार में 20 सेकंड का वीडियो और मूल ऑडियो उत्पन्न करने का समर्थन करता है","summary":"ब्लैक फॉरेस्ट लैब्स ने एर्ली एक्सेस तरीके से FLUX 3 मल्टीमॉडल बेस मॉडल लॉन्च किया, जो एकीकृत आर्किटेक्चर का उपयोग करके इमेज, वीडियो और ऑडियो को संयुक्त रूप से सीखता है। यह मॉडल सेल्फ-फ्लो लर्निंग फ्रेमवर्क पर आधारित है और एक ही बार जनरेशन में अधिकतम 20 सेकंड का वीडियो आउटपुट कर सकता है, जिसमें मूल ऑडियो शामिल होता है, और यह टेक्स्ट-टू-वीडियो, इमेज-टू-वीडियो, मल्टी-कैमरा सीक्वेंसिंग जैसी टास्क को सपोर्ट करता है।","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"ब्लैक फॉरेस्ट लैब्स ने FLUX 3 मल्टीमॉडल मॉडल जारी किया, जो एक बार में 20 सेकंड का वीडियो और मूल ऑडियो उत्पन्न करने का समर्थन करता है - Aioga AI समाचार","description":"ब्लैक फॉरेस्ट लैब्स ने एर्ली एक्सेस तरीके से FLUX 3 मल्टीमॉडल बेस मॉडल लॉन्च किया, जो एकीकृत आर्किटेक्चर का उपयोग करके इमेज, वीडियो और ऑडियो को संयुक्त रूप से सीखता है। यह मॉडल सेल...","url":"https://www.aioga.com/hi/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:28:01.884Z"},"it":{"title":"Black Forest Labs ha rilasciato il modello multimodale FLUX 3, che supporta la generazione di video di 20 secondi e audio nativo in un unico passaggio","summary":"Black Forest Labs ha lanciato il modello di base multimodale FLUX 3 in Accesso Anticipato, utilizzando un'architettura unificata per l'apprendimento congiunto di immagini, video e audio. Questo modello, basato sul framework di apprendimento Self-Flow esteso, può generare video fino a 20 secondi in un'unica produzione con audio nativo, supportando attività come la generazione di video da testo, la generazione di video da immagini e l'assemblaggio di più telecamere.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs ha rilasciato il modello multimodale FLUX 3, che supporta la generazione di video di 20 secondi e audio nativo in un unico passaggio - Aioga Notizie IA","description":"Black Forest Labs ha lanciato il modello di base multimodale FLUX 3 in Accesso Anticipato, utilizzando un'architettura unificata per l'apprendimento congiunto di immagini, video e...","url":"https://www.aioga.com/it/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:28:38.662Z"},"nl":{"title":"Black Forest Labs heeft het FLUX 3 multimodale model uitgebracht, dat het genereren van 20 seconden video en native audio in één keer ondersteunt","summary":"Black Forest Labs heeft het FLUX 3 multimodale basismodel uitgebracht in Early Access, met een uniforme architectuur die het gezamenlijk leren van afbeeldingen, video en audio mogelijk maakt. Dit model is uitgebreid op basis van het Self-Flow leerraamwerk en kan in een enkele generatie video's van maximaal 20 seconden produceren met originele audio, en ondersteunt taken zoals tekst-naar-video, afbeelding-naar-video en multi-shot aaneenschakeling.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs heeft het FLUX 3 multimodale model uitgebracht, dat het genereren van 20 seconden video en native audio in één keer ondersteunt - Aioga AI-nieuws","description":"Black Forest Labs heeft het FLUX 3 multimodale basismodel uitgebracht in Early Access, met een uniforme architectuur die het gezamenlijk leren van afbeeldingen, video en audio moge...","url":"https://www.aioga.com/nl/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:28:38.420Z"},"tr":{"title":"Black Forest Labs, tek seferde 20 saniyelik video ve özgün ses üretimini destekleyen FLUX 3 çok modlu modelini yayınladı","summary":"Black Forest Labs, FLUX 3 çok modlu temel modeli Erken Erişim yöntemiyle sunuyor ve görüntü, video ve sesi birleşik bir mimari ile birlikte öğreniyor. Bu model, Self-Flow öğrenme çerçevesi üzerine genişletilmiş olup, tek seferde en uzun 20 saniyelik video ve orijinal sesi çıkartabilir; metinden video, görüntüden video, çok kamera zincirleme gibi görevleri destekler.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs, tek seferde 20 saniyelik video ve özgün ses üretimini destekleyen FLUX 3 çok modlu modelini yayınladı - Aioga AI Haberleri","description":"Black Forest Labs, FLUX 3 çok modlu temel modeli Erken Erişim yöntemiyle sunuyor ve görüntü, video ve sesi birleşik bir mimari ile birlikte öğreniyor. Bu model, Self-Flow öğrenme ç...","url":"https://www.aioga.com/tr/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:29:17.187Z"},"vi":{"title":"Black Forest Labs phát hành mô hình đa phương thức FLUX 3, hỗ trợ tạo video 20 giây và âm thanh gốc trong một lần","summary":"Black Forest Labs ra mắt mô hình nền tảng đa phương thức FLUX 3 theo hình thức Early Access, sử dụng kiến trúc thống nhất để học kết hợp hình ảnh, video và âm thanh. Mô hình này được mở rộng dựa trên khung học Self-Flow, có thể xuất video dài tối đa 20 giây kèm âm thanh gốc trong một lần tạo, hỗ trợ các nhiệm vụ như tạo video từ văn bản, tạo video từ hình ảnh và ghép nối nhiều cảnh quay.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs phát hành mô hình đa phương thức FLUX 3, hỗ trợ tạo video 20 giây và âm thanh gốc trong một lần - Tin tức AI Aioga","description":"Black Forest Labs ra mắt mô hình nền tảng đa phương thức FLUX 3 theo hình thức Early Access, sử dụng kiến trúc thống nhất để học kết hợp hình ảnh, video và âm thanh. Mô hình này đư...","url":"https://www.aioga.com/vi/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:29:16.896Z"},"id":{"title":"Black Forest Labs merilis model multimodal FLUX 3, mendukung pembuatan video 20 detik sekaligus dengan audio asli","summary":"Black Forest Labs meluncurkan model dasar multimodal FLUX 3 melalui Akses Awal, menggunakan arsitektur terpadu untuk belajar secara bersamaan gambar, video, dan audio. Model ini diperluas berdasarkan kerangka pembelajaran Self-Flow, dapat menghasilkan video sampai 20 detik dengan audio asli dalam satu kali proses, dan mendukung tugas seperti video dari teks, video dari gambar, serta penggabungan multi-kamera.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs merilis model multimodal FLUX 3, mendukung pembuatan video 20 detik sekaligus dengan audio asli - Berita AI Aioga","description":"Black Forest Labs meluncurkan model dasar multimodal FLUX 3 melalui Akses Awal, menggunakan arsitektur terpadu untuk belajar secara bersamaan gambar, video, dan audio. Model ini di...","url":"https://www.aioga.com/id/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:29:54.655Z"},"th":{"title":"Black Forest Labs เปิดตัวโมเดลมัลติโมดอล FLUX 3 รองรับการสร้างวิดีโอ 20 วินาทีพร้อมเสียงต้นฉบับในครั้งเดียว","summary":"Black Forest Labs เปิดตัวโมเดลพื้นฐานหลายโหมด FLUX 3 ในรูปแบบ Early Access โดยใช้สถาปัตยกรรมแบบรวมเพื่อเรียนรู้ร่วมกันทั้งภาพ วิดีโอ และเสียง โมเดลนี้ขยายจากกรอบการเรียนรู้ Self-Flow สามารถสร้างวิดีโอความยาวสูงสุด 20 วินาทีพร้อมเสียงต้นฉบับได้ในการสร้างครั้งเดียว และรองรับงานต่าง ๆ เช่น การสร้างวิดีโอจากข้อความ การสร้างวิดีโอจากภาพ และการเชื่อมต่อหลายมุมกล้อง","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs เปิดตัวโมเดลมัลติโมดอล FLUX 3 รองรับการสร้างวิดีโอ 20 วินาทีพร้อมเสียงต้นฉบับในครั้งเดียว - ข่าว AI Aioga","description":"Black Forest Labs เปิดตัวโมเดลพื้นฐานหลายโหมด FLUX 3 ในรูปแบบ Early Access โดยใช้สถาปัตยกรรมแบบรวมเพื่อเรียนรู้ร่วมกันทั้งภาพ วิดีโอ และเสียง โมเดลนี้ขยายจากกรอบการเรียนรู้ Self-Fl...","url":"https://www.aioga.com/th/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:30:06.355Z"},"pl":{"title":"Black Forest Labs wprowadza wielomodalny model FLUX 3, obsługujący jednokrotne generowanie 20-sekundowego wideo z natywnym dźwiękiem","summary":"Black Forest Labs wprowadza model podstawowy FLUX 3 w trybie wczesnego dostępu, wykorzystujący jednolitą architekturę do wspólnego uczenia się obrazów, wideo i dźwięku. Model oparty jest na rozszerzeniu ramki uczenia Self-Flow, może generować w jednej sesji wideo o długości do 20 sekund z natywnym dźwiękiem i obsługuje zadania takie jak generowanie wideo z tekstu, generowanie wideo z obrazu czy łączenie wielu ujęć.","category":"模型更新","source":"IT之家（RSS）","aggregationSource":"IT之家（RSS）","pageTitle":"Black Forest Labs wprowadza wielomodalny model FLUX 3, obsługujący jednokrotne generowanie 20-sekundowego wideo z natywnym dźwiękiem - Aioga Wiadomości AI","description":"Black Forest Labs wprowadza model podstawowy FLUX 3 w trybie wczesnego dostępu, wykorzystujący jednolitą architekturę do wspólnego uczenia się obrazów, wideo i dźwięku. Model opart...","url":"https://www.aioga.com/pl/news/cmrykywyx026jrolgq06set4j/","contentTranslated":true,"sourceHash":"29df45bd0b3500a2","translatedAt":"2026-07-26T02:30:48.827Z"}}}}