[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-post-\u002Ffr\u002Fnews\u002Fopenai-sora-arret-spud-prochain-flagship-modele-gpt6":3,"news-related-fr-\u002Ffr\u002Fnews\u002Fopenai-sora-arret-spud-prochain-flagship-modele-gpt6":435,"hreflang-openai-sora-shutdown-spud-flagship":455},{"doc":4,"redirect":8},{"id":5,"title":6,"author":7,"author_avatar":8,"body":9,"category":410,"date":411,"description":412,"extension":413,"image":414,"meta":415,"navigation":416,"path":417,"seo":418,"seoDescription":419,"seoTitle":420,"stem":421,"tags":422,"translationKey":433,"updatedAt":8,"__hash__":434},"news_fr\u002Ffr\u002Fnews\u002Fopenai-sora-arret-spud-prochain-flagship-modele-gpt6.md","OpenAI arrête Sora et mise tout sur Spud : le prochain flagship arrive dans quelques semaines — ce qu'on sait déjà","Paul Lefizelier",null,{"type":10,"value":11,"toc":396},"minimark",[12,25,30,33,36,48,51,55,58,61,69,72,76,86,95,98,101,105,108,115,118,122,125,136,144,147,150,155,237,241,345,347,351,388,390],[13,14,15,16,20,21],"p",{},"Trois mois après avoir signé un deal de 3 ans avec Disney, OpenAI vient d'arrêter Sora — son générateur vidéo IA lancé fin 2024. La raison : concentrer toutes les ressources GPU (processeurs graphiques dédiés au calcul IA) sur ",[17,18,19],"strong",{},"Spud",", le prochain flagship model dont le pre-training (phase d'entraînement initial du modèle) est terminé. Sortie annoncée avant fin mai 2026. Sam Altman l'a dit clairement : ",[22,23,24],"em",{},"\"Things are moving faster than many expected.\"",[26,27,29],"h2",{"id":28},"sora-lancé-en-fanfare-arrêté-en-16-mois","Sora : lancé en fanfare, arrêté en 16 mois",[13,31,32],{},"Sora, c'était le pari d'OpenAI sur la vidéo générative. Lancé fin 2024, annoncé comme la révolution de la création vidéo par intelligence artificielle. Le monde entier en parlait.",[13,34,35],{},"En décembre 2025, OpenAI signe un deal majeur avec Disney. Trois ans de partenariat. Des personnages Marvel, Pixar, Star Wars générés par IA sous licence. L'ambition est claire : la vidéo générative devient le prochain grand marché après le texte.",[13,37,38,39,47],{},"Seize mois après le lancement, c'est terminé. Comme le rapporte le ",[40,41,46],"a",{"href":42,"rel":43,"target":45},"https:\u002F\u002Fwww.nytimes.com\u002F2026\u002F03\u002F24\u002Ftechnology\u002Fopenai-shutting-down-sora.html",[44],"nofollow","_blank","New York Times",", OpenAI ferme Sora fin mars 2026. Pas parce que le produit échouait. Mais parce qu'OpenAI a un problème plus urgent à résoudre.",[13,49,50],{},"Le deal Disney ? Rendu caduc. Trois mois après sa signature. C'est le signe que même les mega-deals corporate ne résistent pas aux pivots stratégiques de l'industrie IA en 2026.",[26,52,54],{"id":53},"le-problème-gemini-25-pro-a-changé-le-jeu","Le problème : Gemini 2.5 Pro a changé le jeu",[13,56,57],{},"Mi-mars 2026, Google lance Gemini 2.5 Pro. Le modèle domine LMSYS (le classement de référence qui évalue les LLMs via des votes humains en aveugle). Il écrase les benchmarks (tests standardisés) en coding, en math, en raisonnement.",[13,59,60],{},"En face, GPT-4.5 — sorti en février 2026 — est jugé décevant par rapport aux attentes. Pour la première fois, OpenAI se retrouve clairement derrière Google sur les modèles flagship (les modèles les plus puissants d'un laboratoire).",[13,62,63,64,68],{},"Le calcul est simple. Chaque GPU dédié à Sora est un GPU qui ne sert pas à accélérer Spud. Et dans une course où ",[40,65,67],{"href":66},"\u002Ffr\u002Fnews\u002Fclaude-mythos-fuite-anthropic-modele-dangereux-cybersecurite","Claude Opus 4.6"," d'Anthropic est solide, où Grok 3 de xAI monte en puissance, OpenAI n'a pas le luxe de disperser ses ressources de calcul.",[13,70,71],{},"La vidéo IA peut attendre. La guerre des LLMs flagship, non.",[26,73,75],{"id":74},"spud-ce-que-les-leaks-et-altman-confirment","Spud : ce que les leaks et Altman confirment",[13,77,78,79,82,83,85],{},"Le 24 mars 2026, The Information révèle qu'OpenAI a terminé le pre-training d'un modèle qualifié de ",[22,80,81],{},"\"very strong\"",". Son nom de code interne : ",[17,84,19],{},".",[13,87,88,89,94],{},"Quelques jours plus tard, Sam Altman confirme. ",[40,90,93],{"href":91,"rel":92,"target":45},"https:\u002F\u002Fwww.marketscreener.com\u002Fnews\u002Faltman-said-openai-had-completed-the-initial-development-of-its-next-major-ai-model-codename-spud\u002F",[44],"Sortie prévue avant Memorial Day"," (le dernier lundi de mai aux États-Unis) — soit avant fin mai 2026.",[13,96,97],{},"\"Spud\" comme nom de code s'inscrit dans la tradition OpenAI des noms alimentaires absurdes. \"Strawberry\" pour o1, \"Orion\" pour les versions internes d'o3. Cette fois, c'est une pomme de terre.",[13,99,100],{},"Ce qu'on ne sait pas encore : l'architecture exacte, la taille du modèle, ses capacités multimodales, le pricing, et surtout — si la génération vidéo sera intégrée dans Spud ou si un Sora 2.0 verra le jour après.",[26,102,104],{"id":103},"_730-milliards-de-dollars-de-valorisation-et-spud-doit-justifier-chaque-dollar","730 milliards de dollars de valorisation — et Spud doit justifier chaque dollar",[13,106,107],{},"OpenAI vient de lever 110 milliards de dollars. Valorisation pre-money : 730 milliards de dollars. SoftBank emprunte 40 milliards de dollars spécifiquement pour financer sa participation. Nvidia et Amazon sont dans le tour de table.",[13,109,110,111,114],{},"À ce niveau de capital concentré, chaque trimestre sans flagship compétitif est un risque existentiel. Spud n'est pas juste un modèle. C'est la preuve qu'OpenAI mérite sa valorisation dans un environnement où Google et Anthropic — avec ",[40,112,113],{"href":66},"Claude Mythos en coulisses"," — ne s'endorment pas.",[13,116,117],{},"En parallèle, Anthropic prépare une possible IPO en octobre 2026. La concentration du capital IA autour d'OpenAI n'a jamais été aussi forte. Et OpenAI choisit de couper des produits live pour concentrer ses ressources sur un seul modèle.",[26,119,121],{"id":120},"ce-que-ça-dit-du-rythme-de-lindustrie-ia-en-2026","Ce que ça dit du rythme de l'industrie IA en 2026",[13,123,124],{},"Les deals corporate de 3 ans sont obsolètes avant d'être exécutés. Les produits lancés en grande pompe sont arrêtés en moins de 2 ans si la stratégie pivote. Bienvenue en 2026.",[13,126,127,128,132,133,135],{},"Jensen Huang, CEO de Nvidia, a dit lors du ",[40,129,131],{"href":130},"\u002Ffr\u002Fnews\u002Fnvidia-gtc-2026-keynote-jensen-huang","GTC 2026"," : ",[22,134,24],{}," Sam Altman a repris exactement la même formulation. Ce n'est pas un hasard. C'est la nouvelle réalité de l'industrie.",[13,137,138,139,143],{},"L'accélération est partout. ",[40,140,142],{"href":141},"\u002Ffr\u002Fnews\u002Fmeta-hyperagents-self-modification-metacognitive-apprentissage-autonome","Meta publie des agents qui réécrivent leur propre code",". Anthropic construit des modèles trop dangereux pour être distribués. Google domine les benchmarks avec Gemini 2.5 Pro. Et OpenAI sacrifie un deal Disney pour libérer des GPU.",[13,145,146],{},"Le rythme ne ralentit pas. Il accélère.",[148,149],"hr",{},[151,152,154],"h3",{"id":153},"la-timeline-des-modèles-openai-2025-2026","La timeline des modèles OpenAI (2025-2026)",[156,157,158,174],"table",{},[159,160,161],"thead",{},[162,163,164,168,171],"tr",{},[165,166,167],"th",{},"Modèle",[165,169,170],{},"Date",[165,172,173],{},"Verdict",[175,176,177,189,199,210,221],"tbody",{},[162,178,179,183,186],{},[180,181,182],"td",{},"GPT-4o (maj)",[180,184,185],{},"2025",[180,187,188],{},"✅ Solide",[162,190,191,194,196],{},[180,192,193],{},"o1 \u002F o3",[180,195,185],{},[180,197,198],{},"✅ Raisonnement",[162,200,201,204,207],{},[180,202,203],{},"GPT-4.5",[180,205,206],{},"Février 2026",[180,208,209],{},"⚠️ Décevant vs attentes",[162,211,212,215,218],{},[180,213,214],{},"Sora",[180,216,217],{},"2024 → Mars 2026",[180,219,220],{},"❌ Arrêté",[162,222,223,227,232],{},[180,224,225],{},[17,226,19],{},[180,228,229],{},[17,230,231],{},"Avant fin mai 2026",[180,233,234],{},[17,235,236],{},"🔜 Pre-training terminé",[151,238,240],{"id":239},"la-guerre-des-flagship-llms-mars-2026","La guerre des flagship LLMs (mars 2026)",[156,242,243,258],{},[159,244,245],{},[162,246,247,249,252,255],{},[165,248,167],{},[165,250,251],{},"Labo",[165,253,254],{},"Statut",[165,256,257],{},"Point fort",[175,259,260,274,286,299,312,325],{},[162,261,262,265,268,271],{},[180,263,264],{},"Gemini 2.5 Pro",[180,266,267],{},"Google",[180,269,270],{},"✅ Disponible",[180,272,273],{},"LMSYS n°1, coding, math",[162,275,276,278,281,283],{},[180,277,67],{},[180,279,280],{},"Anthropic",[180,282,270],{},[180,284,285],{},"Raisonnement, safety",[162,287,288,291,293,296],{},[180,289,290],{},"Claude Mythos",[180,292,280],{},[180,294,295],{},"🔒 Early access restreint",[180,297,298],{},"Cyber, code",[162,300,301,304,307,309],{},[180,302,303],{},"Grok 3",[180,305,306],{},"xAI",[180,308,270],{},[180,310,311],{},"Raisonnement, vitesse",[162,313,314,316,319,322],{},[180,315,203],{},[180,317,318],{},"OpenAI",[180,320,321],{},"⚠️ Décevant",[180,323,324],{},"Créativité",[162,326,327,331,335,340],{},[180,328,329],{},[17,330,19],{},[180,332,333],{},[17,334,318],{},[180,336,337],{},[17,338,339],{},"🔜 Semaines",[180,341,342],{},[17,343,344],{},"\"Very strong\" (Altman)",[148,346],{},[26,348,350],{"id":349},"en-résumé","En résumé",[352,353,354,361,370,376,382],"ul",{},[355,356,357,360],"li",{},[17,358,359],{},"OpenAI arrête Sora fin mars 2026",", 16 mois après son lancement et 3 mois après un deal de 3 ans avec Disney (Marvel, Pixar, Star Wars), pour concentrer ses ressources GPU sur Spud",[355,362,363,366,367,369],{},[17,364,365],{},"Spud est le prochain flagship model d'OpenAI"," : pre-training terminé, qualifié de ",[22,368,81],{},", sortie annoncée avant fin mai 2026 (avant Memorial Day)",[355,371,372,375],{},[17,373,374],{},"Sam Altman : \"Things are moving faster than many expected\""," — Spud est la réponse directe à Gemini 2.5 Pro de Google, qui domine les benchmarks depuis mi-mars",[355,377,378,381],{},[17,379,380],{},"OpenAI vient de lever 110 milliards de dollars"," à 730 milliards de valorisation, avec SoftBank empruntant 40 milliards pour financer sa participation",[355,383,384,387],{},[17,385,386],{},"Signal industrie"," : en 2026, les deals corporate de 3 ans et les produits lancés en fanfare cèdent la place aux pivots stratégiques en quelques semaines",[148,389],{},[13,391,392,393,395],{},"Sora n'était pas un échec. C'était une distraction. Et OpenAI vient de choisir entre deux guerres : celle de la vidéo générative et celle des LLMs flagship. Ils ont choisi la deuxième — parce que c'est là que se joue la valorisation à 730 milliards de dollars. Spud arrive dans quelques semaines. Google a Gemini 2.5 Pro. Anthropic a Mythos en coulisses. Et OpenAI a sacrifié un deal Disney pour avoir suffisamment de GPU pour répondre. ",[22,394,24],{}," C'est la phrase la plus honnête de l'industrie IA depuis le début de l'année.",{"title":397,"searchDepth":398,"depth":398,"links":399},"",2,[400,401,402,403,404,409],{"id":28,"depth":398,"text":29},{"id":53,"depth":398,"text":54},{"id":74,"depth":398,"text":75},{"id":103,"depth":398,"text":104},{"id":120,"depth":398,"text":121,"children":405},[406,408],{"id":153,"depth":407,"text":154},3,{"id":239,"depth":407,"text":240},{"id":349,"depth":398,"text":350},"ai","2026-03-31T09:05:00","OpenAI arrête Sora trois mois après un deal Disney et concentre ses GPU sur Spud, son prochain flagship model. Pre-training terminé, sortie avant fin mai 2026.","md","\u002Fopenai-sora-arret-spud-prochain-flagship-modele-gpt6.jpg",{},true,"\u002Ffr\u002Fnews\u002Fopenai-sora-arret-spud-prochain-flagship-modele-gpt6",{"title":6,"description":412},"OpenAI arrête Sora et concentre ses GPU sur Spud, prochain flagship model. Pre-training terminé, sortie avant fin mai 2026. Deal Disney annulé, 730 Mds$ de valorisation.","OpenAI arrête Sora, mise sur Spud : flagship model imminent","fr\u002Fnews\u002Fopenai-sora-arret-spud-prochain-flagship-modele-gpt6",[423,424,425,426,427,428,429,430,431,432],"openai","sora","spud","gpt6","sam-altman","flagship-model","video-generation","disney","softbank","pre-training","openai-sora-shutdown-spud-flagship","AUkwcNBdxB9xparvLJsR-c8rLtEgCzQsO2ETOsw8aO8",[436,442,449],{"title":437,"description":438,"path":439,"date":440,"image":441,"category":410},"Meta suspend ses nouvelles conditions publicitaires après la fronde contre le scraping pour l'IA","Meta a suspendu sans date la mise à jour de quatre jeux de conditions commerciales et publicitaires qui, selon les éditeurs, lui auraient permis d'aspirer leurs sites pour ses produits d'IA et de publicité.","\u002Ffr\u002Fnews\u002Fmeta-suspend-conditions-publicitaires-scraping-ia","2026-11-09T08:00:00","\u002Fimages\u002Fnews\u002Fmeta-pauses-business-ad-terms-ai-scraping.jpg",{"title":443,"description":444,"path":445,"date":446,"image":447,"category":448},"L'option « ne plus demander » de Claude Code a laissé des mots de passe dans 1 122 fichiers publics sur GitHub","La société de sécurité Apiiro a trouvé 1 122 copies publiques du fichier de réglages locaux de Claude Code contenant des secrets, après que « Yes, and don't ask again » a enregistré des commandes approuvées en clair.","\u002Ffr\u002Fnews\u002Fclaude-code-ne-plus-demander-secrets-github","2026-11-08T08:00:00","\u002Fimages\u002Fnews\u002Fclaude-code-dont-ask-again-passwords-github.jpg","vibecoding",{"title":450,"description":451,"path":452,"date":453,"image":454,"category":448},"Les cloud agents Cursor construisent, hébergent et publient des apps live","Les cloud agents Cursor peuvent scaffolder une app web, provisionner Postgres et le SSO Cursor, lancer des contrôles de sécurité, puis publier une URL partageable — sans compte d'hébergement séparé.","\u002Ffr\u002Fnews\u002Fcursor-live-apps-agents-cloud-deploiement-postgres-auth","2026-11-07T08:00:00","\u002Fimages\u002Fnews\u002Fcursor-live-apps-cloud-agents-deploy-postgres-auth.jpg",{"en":456,"fr":827,"es":8},{"id":457,"title":458,"author":7,"author_avatar":8,"body":459,"category":410,"date":411,"description":818,"extension":413,"image":414,"meta":819,"navigation":416,"path":820,"seo":821,"seoDescription":822,"seoTitle":823,"stem":824,"tags":825,"translationKey":433,"updatedAt":8,"__hash__":826},"news_en\u002Fnews\u002Fopenai-sora-shutdown-spud-next-flagship-model-gpt6.md","OpenAI Shuts Down Sora and Bets Everything on Spud: The Next Flagship Model Is Weeks Away — What We Know",{"type":10,"value":460,"toc":807},[461,469,473,476,479,487,490,494,497,500,507,510,514,523,531,534,537,541,544,551,554,558,561,571,579,582,584,588,658,662,757,759,763,799,801],[13,462,463,464,466,467],{},"Three months after signing a 3-year deal with Disney, OpenAI just shut down Sora — its AI video generator launched in late 2024. The reason: redirecting all GPU resources (graphics processing units dedicated to AI computation) toward ",[17,465,19],{},", the next flagship model whose pre-training (the initial training phase of an AI model) is complete. Launch expected before the end of May 2026. Sam Altman said it plainly: ",[22,468,24],{},[26,470,472],{"id":471},"sora-launched-with-fanfare-killed-in-16-months","Sora: Launched With Fanfare, Killed in 16 Months",[13,474,475],{},"Sora was OpenAI's bet on generative video. Launched in late 2024, hailed as the revolution in AI-powered video creation. The whole world was watching.",[13,477,478],{},"In December 2025, OpenAI signed a major deal with Disney. Three years of partnership. Marvel, Pixar, and Star Wars characters generated by AI under license. The ambition was clear: generative video would become the next major market after text.",[13,480,481,482,486],{},"Sixteen months after launch, it's over. As ",[40,483,485],{"href":42,"rel":484,"target":45},[44],"reported by the New York Times",", OpenAI is shutting down Sora at the end of March 2026. Not because the product was failing. But because OpenAI has a more urgent problem to solve.",[13,488,489],{},"The Disney deal? Voided. Three months after it was signed. It's a sign that even mega corporate partnerships can't survive the strategic pivots of the AI industry in 2026.",[26,491,493],{"id":492},"the-problem-gemini-25-pro-changed-the-game","The Problem: Gemini 2.5 Pro Changed the Game",[13,495,496],{},"In mid-March 2026, Google launched Gemini 2.5 Pro. The model dominates LMSYS (the reference leaderboard that evaluates LLMs through blind human voting). It crushes benchmarks (standardized tests) in coding, math, and reasoning.",[13,498,499],{},"Meanwhile, GPT-4.5 — released in February 2026 — was judged disappointing relative to expectations. For the first time, OpenAI finds itself clearly behind Google on flagship models (a lab's most powerful models).",[13,501,502,503,506],{},"The math is simple. Every GPU dedicated to Sora is a GPU not being used to accelerate Spud. And in a race where Anthropic's ",[40,504,67],{"href":505},"\u002Fnews\u002Fclaude-mythos-leak-anthropic-most-powerful-model-cybersecurity"," is strong, where xAI's Grok 3 is gaining ground, OpenAI can't afford to spread its compute resources thin.",[13,508,509],{},"AI video can wait. The flagship LLM war cannot.",[26,511,513],{"id":512},"spud-what-leaks-and-altman-confirm","Spud: What Leaks and Altman Confirm",[13,515,516,517,520,521,85],{},"On March 24, 2026, The Information reported that OpenAI completed pre-training on a model described as ",[22,518,519],{},"\"very strong.\""," Its internal codename: ",[17,522,19],{},[13,524,525,526,530],{},"Days later, Sam Altman confirmed. ",[40,527,529],{"href":91,"rel":528,"target":45},[44],"Launch expected before Memorial Day"," (the last Monday of May in the United States) — meaning before the end of May 2026.",[13,532,533],{},"\"Spud\" as a codename fits OpenAI's tradition of absurd food-related names. \"Strawberry\" for o1, \"Orion\" for internal versions of o3. This time, it's a potato.",[13,535,536],{},"What we don't know yet: the exact architecture, model size, multimodal capabilities, pricing, and crucially — whether video generation will be integrated into Spud or whether a Sora 2.0 will emerge afterward.",[26,538,540],{"id":539},"_730-billion-valuation-and-spud-must-justify-every-dollar","$730 Billion Valuation — and Spud Must Justify Every Dollar",[13,542,543],{},"OpenAI just raised $110 billion. Pre-money valuation: $730 billion. SoftBank is borrowing $40 billion specifically to fund its share. Nvidia and Amazon are in the round.",[13,545,546,547,550],{},"At this level of concentrated capital, every quarter without a competitive flagship is an existential risk. Spud isn't just a model. It's proof that OpenAI deserves its valuation in an environment where Google and Anthropic — with ",[40,548,549],{"href":505},"Claude Mythos behind the scenes"," — aren't standing still.",[13,552,553],{},"Meanwhile, Anthropic is preparing a potential IPO in October 2026. The concentration of AI capital around OpenAI has never been this intense. And OpenAI is choosing to cut live products to focus resources on a single model.",[26,555,557],{"id":556},"what-this-says-about-the-pace-of-ai-in-2026","What This Says About the Pace of AI in 2026",[13,559,560],{},"Three-year corporate deals become obsolete before they're executed. Products launched with fanfare get killed in under two years when strategy pivots. Welcome to 2026.",[13,562,563,564,567,568,570],{},"Jensen Huang, Nvidia's CEO, said at ",[40,565,131],{"href":566},"\u002Fnews\u002Fnvidia-gtc-2026-keynote-jensen-huang",": ",[22,569,24],{}," Sam Altman used the exact same phrasing. That's not a coincidence. It's the new reality of the industry.",[13,572,573,574,578],{},"The acceleration is everywhere. ",[40,575,577],{"href":576},"\u002Fnews\u002Fmeta-hyperagents-self-modification-metacognitive-autonomous-learning","Meta publishes agents that rewrite their own code",". Anthropic builds models too dangerous to release. Google dominates benchmarks with Gemini 2.5 Pro. And OpenAI sacrifices a Disney deal to free up GPU compute.",[13,580,581],{},"The pace isn't slowing down. It's accelerating.",[148,583],{},[151,585,587],{"id":586},"openai-model-timeline-2025-2026","OpenAI Model Timeline (2025-2026)",[156,589,590,601],{},[159,591,592],{},[162,593,594,597,599],{},[165,595,596],{},"Model",[165,598,170],{},[165,600,173],{},[175,602,603,613,622,632,642],{},[162,604,605,608,610],{},[180,606,607],{},"GPT-4o (update)",[180,609,185],{},[180,611,612],{},"✅ Solid",[162,614,615,617,619],{},[180,616,193],{},[180,618,185],{},[180,620,621],{},"✅ Reasoning",[162,623,624,626,629],{},[180,625,203],{},[180,627,628],{},"February 2026",[180,630,631],{},"⚠️ Disappointing vs expectations",[162,633,634,636,639],{},[180,635,214],{},[180,637,638],{},"2024 → March 2026",[180,640,641],{},"❌ Shut down",[162,643,644,648,653],{},[180,645,646],{},[17,647,19],{},[180,649,650],{},[17,651,652],{},"Before end of May 2026",[180,654,655],{},[17,656,657],{},"🔜 Pre-training complete",[151,659,661],{"id":660},"the-flagship-llm-war-march-2026","The Flagship LLM War (March 2026)",[156,663,664,679],{},[159,665,666],{},[162,667,668,670,673,676],{},[165,669,596],{},[165,671,672],{},"Lab",[165,674,675],{},"Status",[165,677,678],{},"Strength",[175,680,681,693,704,715,726,738],{},[162,682,683,685,687,690],{},[180,684,264],{},[180,686,267],{},[180,688,689],{},"✅ Available",[180,691,692],{},"LMSYS #1, coding, math",[162,694,695,697,699,701],{},[180,696,67],{},[180,698,280],{},[180,700,689],{},[180,702,703],{},"Reasoning, safety",[162,705,706,708,710,713],{},[180,707,290],{},[180,709,280],{},[180,711,712],{},"🔒 Restricted early access",[180,714,298],{},[162,716,717,719,721,723],{},[180,718,303],{},[180,720,306],{},[180,722,689],{},[180,724,725],{},"Reasoning, speed",[162,727,728,730,732,735],{},[180,729,203],{},[180,731,318],{},[180,733,734],{},"⚠️ Disappointing",[180,736,737],{},"Creativity",[162,739,740,744,748,753],{},[180,741,742],{},[17,743,19],{},[180,745,746],{},[17,747,318],{},[180,749,750],{},[17,751,752],{},"🔜 Weeks away",[180,754,755],{},[17,756,344],{},[148,758],{},[26,760,762],{"id":761},"key-takeaways","Key Takeaways",[352,764,765,771,781,787,793],{},[355,766,767,770],{},[17,768,769],{},"OpenAI shuts down Sora in late March 2026",", 16 months after launch and 3 months after a 3-year Disney deal (Marvel, Pixar, Star Wars), to concentrate all GPU resources on Spud",[355,772,773,776,777,780],{},[17,774,775],{},"Spud is OpenAI's next flagship model",": pre-training complete, described as ",[22,778,779],{},"\"very strong,\""," launch announced before end of May 2026 (before Memorial Day)",[355,782,783,786],{},[17,784,785],{},"Sam Altman: \"Things are moving faster than many expected\""," — Spud is the direct response to Google's Gemini 2.5 Pro, which has dominated benchmarks since mid-March",[355,788,789,792],{},[17,790,791],{},"OpenAI just raised $110 billion"," at a $730 billion valuation, with SoftBank borrowing $40 billion to fund its stake",[355,794,795,798],{},[17,796,797],{},"Industry signal",": in 2026, three-year corporate deals and flagship product launches give way to strategic pivots executed in weeks",[148,800],{},[13,802,803,804,806],{},"Sora wasn't a failure. It was a distraction. And OpenAI just chose between two wars: the generative video war and the flagship LLM war. They chose the second — because that's where the $730 billion valuation is at stake. Spud arrives in a few weeks. Google has Gemini 2.5 Pro. Anthropic has Mythos in the wings. And OpenAI sacrificed a Disney deal to have enough GPU to respond. ",[22,805,24],{}," It's the most honest sentence the AI industry has produced all year.",{"title":397,"searchDepth":398,"depth":398,"links":808},[809,810,811,812,813,817],{"id":471,"depth":398,"text":472},{"id":492,"depth":398,"text":493},{"id":512,"depth":398,"text":513},{"id":539,"depth":398,"text":540},{"id":556,"depth":398,"text":557,"children":814},[815,816],{"id":586,"depth":407,"text":587},{"id":660,"depth":407,"text":661},{"id":761,"depth":398,"text":762},"OpenAI kills Sora three months after a Disney deal to focus all GPU resources on Spud, its next flagship model. Pre-training complete, launch before end of May 2026.",{},"\u002Fnews\u002Fopenai-sora-shutdown-spud-next-flagship-model-gpt6",{"title":458,"description":818},"OpenAI shuts down Sora to focus all GPU on Spud, its next flagship model. Pre-training complete, launch before May 2026. Disney deal voided, $730B valuation.","OpenAI Shuts Down Sora, Bets on Spud: Flagship Model Imminent","news\u002Fopenai-sora-shutdown-spud-next-flagship-model-gpt6",[423,424,425,426,427,428,429,430,431,432],"7pdreZgqh9TH-yPeO8reTq8vALpXGx33kcVCDVnK764",{"id":5,"title":6,"author":7,"author_avatar":8,"body":828,"category":410,"date":411,"description":412,"extension":413,"image":414,"meta":1099,"navigation":416,"path":417,"seo":1100,"seoDescription":419,"seoTitle":420,"stem":421,"tags":1101,"translationKey":433,"updatedAt":8,"__hash__":434},{"type":10,"value":829,"toc":1088},[830,836,838,840,842,847,849,851,853,855,859,861,863,869,874,876,878,880,882,886,888,890,892,898,902,904,906,908,968,970,1054,1056,1058,1082,1084],[13,831,15,832,20,834],{},[17,833,19],{},[22,835,24],{},[26,837,29],{"id":28},[13,839,32],{},[13,841,35],{},[13,843,38,844,47],{},[40,845,46],{"href":42,"rel":846,"target":45},[44],[13,848,50],{},[26,850,54],{"id":53},[13,852,57],{},[13,854,60],{},[13,856,63,857,68],{},[40,858,67],{"href":66},[13,860,71],{},[26,862,75],{"id":74},[13,864,78,865,82,867,85],{},[22,866,81],{},[17,868,19],{},[13,870,88,871,94],{},[40,872,93],{"href":91,"rel":873,"target":45},[44],[13,875,97],{},[13,877,100],{},[26,879,104],{"id":103},[13,881,107],{},[13,883,110,884,114],{},[40,885,113],{"href":66},[13,887,117],{},[26,889,121],{"id":120},[13,891,124],{},[13,893,127,894,132,896,135],{},[40,895,131],{"href":130},[22,897,24],{},[13,899,138,900,143],{},[40,901,142],{"href":141},[13,903,146],{},[148,905],{},[151,907,154],{"id":153},[156,909,910,920],{},[159,911,912],{},[162,913,914,916,918],{},[165,915,167],{},[165,917,170],{},[165,919,173],{},[175,921,922,930,938,946,954],{},[162,923,924,926,928],{},[180,925,182],{},[180,927,185],{},[180,929,188],{},[162,931,932,934,936],{},[180,933,193],{},[180,935,185],{},[180,937,198],{},[162,939,940,942,944],{},[180,941,203],{},[180,943,206],{},[180,945,209],{},[162,947,948,950,952],{},[180,949,214],{},[180,951,217],{},[180,953,220],{},[162,955,956,960,964],{},[180,957,958],{},[17,959,19],{},[180,961,962],{},[17,963,231],{},[180,965,966],{},[17,967,236],{},[151,969,240],{"id":239},[156,971,972,984],{},[159,973,974],{},[162,975,976,978,980,982],{},[165,977,167],{},[165,979,251],{},[165,981,254],{},[165,983,257],{},[175,985,986,996,1006,1016,1026,1036],{},[162,987,988,990,992,994],{},[180,989,264],{},[180,991,267],{},[180,993,270],{},[180,995,273],{},[162,997,998,1000,1002,1004],{},[180,999,67],{},[180,1001,280],{},[180,1003,270],{},[180,1005,285],{},[162,1007,1008,1010,1012,1014],{},[180,1009,290],{},[180,1011,280],{},[180,1013,295],{},[180,1015,298],{},[162,1017,1018,1020,1022,1024],{},[180,1019,303],{},[180,1021,306],{},[180,1023,270],{},[180,1025,311],{},[162,1027,1028,1030,1032,1034],{},[180,1029,203],{},[180,1031,318],{},[180,1033,321],{},[180,1035,324],{},[162,1037,1038,1042,1046,1050],{},[180,1039,1040],{},[17,1041,19],{},[180,1043,1044],{},[17,1045,318],{},[180,1047,1048],{},[17,1049,339],{},[180,1051,1052],{},[17,1053,344],{},[148,1055],{},[26,1057,350],{"id":349},[352,1059,1060,1064,1070,1074,1078],{},[355,1061,1062,360],{},[17,1063,359],{},[355,1065,1066,366,1068,369],{},[17,1067,365],{},[22,1069,81],{},[355,1071,1072,375],{},[17,1073,374],{},[355,1075,1076,381],{},[17,1077,380],{},[355,1079,1080,387],{},[17,1081,386],{},[148,1083],{},[13,1085,392,1086,395],{},[22,1087,24],{},{"title":397,"searchDepth":398,"depth":398,"links":1089},[1090,1091,1092,1093,1094,1098],{"id":28,"depth":398,"text":29},{"id":53,"depth":398,"text":54},{"id":74,"depth":398,"text":75},{"id":103,"depth":398,"text":104},{"id":120,"depth":398,"text":121,"children":1095},[1096,1097],{"id":153,"depth":407,"text":154},{"id":239,"depth":407,"text":240},{"id":349,"depth":398,"text":350},{},{"title":6,"description":412},[423,424,425,426,427,428,429,430,431,432]]