#opus48 — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #opus48, aggregated by home.social.
-
Meta veröffentlicht das Modell Muse Spark 1.1 für autonome Agenten-Aufgaben und Computernutzung.
Das System erreicht Leistungswerte von Opus 4.8 bei Betriebskosten von 1,25 USD pro Million Input-Tokens. Der interne Sicherheitsbericht weist das Modell aufgrund verbesserter Hacking-Fähigkeiten als potenzielles Hochrisiko-Modell aus, weshalb nachgeschaltete Filter eingesetzt werden.
-
Meta veröffentlicht das Modell Muse Spark 1.1 für autonome Agenten-Aufgaben und Computernutzung.
Das System erreicht Leistungswerte von Opus 4.8 bei Betriebskosten von 1,25 USD pro Million Input-Tokens. Der interne Sicherheitsbericht weist das Modell aufgrund verbesserter Hacking-Fähigkeiten als potenzielles Hochrisiko-Modell aus, weshalb nachgeschaltete Filter eingesetzt werden.
-
GPT-5.6、7月7日発表か——Claude Fable 5の「空白期間」を狙い撃つ戦略的タイミング、コード流出で3つのサブモデルが判明 — BigGo ファイナンス https://www.yayafa.com/2836239/ #AgenticAi #AI #Anthropic #ArtificialGeneralIntelligence #ArtificialIntelligence #ClaudeFable5 #CODEX #DeepSeekR1 #Gemini35Pro #GPT56 #GPT56Luna #GPT56Sol #GPT56Terra #OpenAI #Opus48 #SolUltra #エージェント型AI #トークン・アポカリプス #人工知能 #汎用人工知能
-
GPT-5.6、7月7日発表か——Claude Fable 5の「空白期間」を狙い撃つ戦略的タイミング、コード流出で3つのサブモデルが判明 — BigGo ファイナンス https://www.yayafa.com/2836239/ #AgenticAi #AI #Anthropic #ArtificialGeneralIntelligence #ArtificialIntelligence #ClaudeFable5 #CODEX #DeepSeekR1 #Gemini35Pro #GPT56 #GPT56Luna #GPT56Sol #GPT56Terra #OpenAI #Opus48 #SolUltra #エージェント型AI #トークン・アポカリプス #人工知能 #汎用人工知能
-
Claude Fable 5が規制解除で復活：性能は健在も、過剰な安全審査に開発者から不満噴出 — BigGo ファイナンス https://www.yayafa.com/2835733/ #AgenticAi #AI #Amazon #Anthropic #AnthropicClaude #ArtificialGeneralIntelligence #ArtificialIntelligence #claude #ClaudeFable5 #Google #Microsoft #Mythosシリーズ #Opus48 #Sonnet5 #エージェント型AI #人工知能 #利用クレジット #安全分類器 #汎用人工知能 #米国輸出規制
-
Claude Fable 5が規制解除で復活：性能は健在も、過剰な安全審査に開発者から不満噴出 — BigGo ファイナンス https://www.yayafa.com/2835733/ #AgenticAi #AI #Amazon #Anthropic #AnthropicClaude #ArtificialGeneralIntelligence #ArtificialIntelligence #claude #ClaudeFable5 #Google #Microsoft #Mythosシリーズ #Opus48 #Sonnet5 #エージェント型AI #人工知能 #利用クレジット #安全分類器 #汎用人工知能 #米国輸出規制
-
Fable 5復活に開発者激怒：Anthropicの最上位モデルに「お前には不相応」判定、密かな格下げ機能が発覚 — BigGo ファイナンス https://www.yayafa.com/2835199/ #a16z #AgenticAi #AI #Anthropic #ArtificialGeneralIntelligence #ArtificialIntelligence #ClaudeCode #DeepSeek #Fable5 #mythos #OpenAI #Opus48 #エージェント型AI #ダリオ・アモデイ #タリク・シヒパル #人工知能 #汎用人工知能
-
Fable 5復活に開発者激怒：Anthropicの最上位モデルに「お前には不相応」判定、密かな格下げ機能が発覚 — BigGo ファイナンス https://www.yayafa.com/2835199/ #a16z #AgenticAi #AI #Anthropic #ArtificialGeneralIntelligence #ArtificialIntelligence #ClaudeCode #DeepSeek #Fable5 #mythos #OpenAI #Opus48 #エージェント型AI #ダリオ・アモデイ #タリク・シヒパル #人工知能 #汎用人工知能
-
Does anyone else see instruction-adherence degrade in long Opus 4.8 & 4.7 sessions despite the instructions being present/re-injected. Is it prompt-structure or model behaviour?
-
Does anyone else see instruction-adherence degrade in long Opus 4.8 & 4.7 sessions despite the instructions being present/re-injected. Is it prompt-structure or model behaviour?
-
RT @GaelBreton: Wenn du mit Opus 4.8 und GPT 5.5 nichts Wertvolles entwickeln kannst, wird dich Fable nicht retten.
mehr auf Arint.info
#Fable #GPT55 #KIEntwicklung #Opus48 #Softwarebau #TechTools #arint_info
-
RT @GaelBreton: Wenn du mit Opus 4.8 und GPT 5.5 nichts Wertvolles entwickeln kannst, wird dich Fable nicht retten.
mehr auf Arint.info
#AIEntwicklung #Fable #GPT55 #KünstlicheIntelligenz #Opus48 #TechNews #arint_info
-
RT @GaelBreton: Wenn du mit Opus 4.8 und GPT 5.5 nichts Wertvolles entwickeln kannst, wird dich Fable nicht retten.
mehr auf Arint.info
#AIEntwicklung #Fable #GPT55 #KünstlicheIntelligenz #Opus48 #TechNews #arint_info
-
Anthropic、企業向けAIサブスクでOpenAIを初めて逆転──米政権が最新モデル公開停止を命じる中で — BigGo ファイナンス https://www.yayafa.com/2825041/ #AgenticAi #AI #Anthropic #AraKharazian #ArtificialGeneralIntelligence #ArtificialIntelligence #ChatGPT #claude #Fable5 #Mythos5 #OpenAI #Opus48 #ramp #SensorTower #エージェント型AI #トランプ米政権 #人工知能 #憲法AI #汎用人工知能 #米国防総省
-
RT @TobiasGloeckler: Interessant, wie gut GLM-5.2 im Frontend-Design sein soll. Bisher habe ich immer Opus-4.6 bzw. jetzt Opus-4.8 verwendet.
mehr auf Arint.info
#FrontendDesign #GLM52 #Opus48 #TechComparison #WebDevelopment #arint_info
-
CW: AI LLM anthropic
Apropos #Fable5: Ich hatte diese Woche das Modell auf Kosten meines Arbeitgebers ausprobiert.
Nachdem #Opus47 und #Opus48 gegenüber #Opus46 eher wie ein Rückschritt anfühlten, scheint Fable 5 wieder auf dem Niveau von Opus 4.6 gewesen zu sein.
Wenig verwunderlich, denn Opus 4.7 und Opus 4.8 sind wesentlich kleiner als Opus 4.6. Fable 5 und #Mythos sollen wohl in etwa so groß wie Opus 4.6 sein.
Ich sehe die Neubenennung daher eher als Versuch an, die Preise zu verdoppeln. 🤷♀️
-
CW: AI LLM anthropic
Apropos #Fable5: Ich hatte diese Woche das Modell auf Kosten meines Arbeitgebers ausprobiert.
Nachdem #Opus47 und #Opus48 gegenüber #Opus46 eher wie ein Rückschritt anfühlten, scheint Fable 5 wieder auf dem Niveau von Opus 4.6 gewesen zu sein.
Wenig verwunderlich, denn Opus 4.7 und Opus 4.8 sind wesentlich kleiner als Opus 4.6. Fable 5 und #Mythos sollen wohl in etwa so groß wie Opus 4.6 sein.
Ich sehe die Neubenennung daher eher als Versuch an, die Preise zu verdoppeln. 🤷♀️
-
The LLM with the lowest hallucination rate is an #openweight model: #MinimaxM3, which was released a few days ago. Hey almighty #Opus48, where are you?!
-
Claude Opus 4.8 has quite a few annoying quirks, and it's important to be precise about which ones, because this is the typical claim that needs to be substantiated by facts rather than hinted at. This is not just uncanny — it's painful to look at.
-
Claude Opus 4.8 has quite a few annoying quirks, and it's important to be precise about which ones, because this is the typical claim that needs to be substantiated by facts rather than hinted at. This is not just uncanny — it's painful to look at.
-
Claude Codeの「Plan mode」を捨てる選択──grill-meとADRで設計のブレを抑える新潮流 — BigGo ファイナンス https://www.yayafa.com/2814941/ #AdrManagerスキル #ADR（ArchitectureDecisionRecords） #AgenticAi #AI #Anthropic #ArtificialGeneralIntelligence #ArtificialIntelligence #ClaudeCode #ClaudeDesign #Fixer #GrillMeスキル #Haiku45 #MagicLink #Opus47 #Opus48 #PlanMode #Sonnet46 #SupabaseAuth #エージェント型AI #人工知能 #汎用人工知能
-
Claude Codeの「Plan mode」を捨てる選択──grill-meとADRで設計のブレを抑える新潮流 — BigGo ファイナンス https://www.yayafa.com/2814941/ #AdrManagerスキル #ADR（ArchitectureDecisionRecords） #AgenticAi #AI #Anthropic #ArtificialGeneralIntelligence #ArtificialIntelligence #ClaudeCode #ClaudeDesign #Fixer #GrillMeスキル #Haiku45 #MagicLink #Opus47 #Opus48 #PlanMode #Sonnet46 #SupabaseAuth #エージェント型AI #人工知能 #汎用人工知能
-
@scottgal I have been getting "that's textbook" analysis from #opus48 even after explicitly ruling "don't guess, verify" but I'm working different projects across sperate dedicated machines per project and then journaling to my homelab topology repo synced everywhere. It might just need to catch up on the drift. But it feels like a short circuit/shortcut in thinking responses. Will find out tomorrow.
-
@scottgal I have been getting "that's textbook" analysis from #opus48 even after explicitly ruling "don't guess, verify" but I'm working different projects across sperate dedicated machines per project and then journaling to my homelab topology repo synced everywhere. It might just need to catch up on the drift. But it feels like a short circuit/shortcut in thinking responses. Will find out tomorrow.
-
Push back on #opus48 if something doesn’t smell right or it’s using the term “textbook”. Make #agents verify their claims by either using web search or tailing logs 🪵 or whatever’s the appropriate way for the situation. #claude will say you’re right and then do the needful. Also TDD everything you can.
-
Claude Opus 4.8 marks a crucial shift from sequential text processing to dynamic, multi-agent orchestration.
While raw benchmarks show steady optimization over 4.7, the real breakthrough lies in architecture: the "dynamic workflows" feature manages context isolation by spawning separate subagents, preventing context memory inflation during massive coding migrations. Furthermore, internal evaluations indicate the model is 4x less likely to overlook flaws in its own code.
-
Claude Opus 4.8 marks a crucial shift from sequential text processing to dynamic, multi-agent orchestration.
While raw benchmarks show steady optimization over 4.7, the real breakthrough lies in architecture: the "dynamic workflows" feature manages context isolation by spawning separate subagents, preventing context memory inflation during massive coding migrations. Furthermore, internal evaluations indicate the model is 4x less likely to overlook flaws in its own code.
-
I am setting the brand new #Opus48 on my main #vibecode project. In the first instance I am getting it to do a project review...
...already super impressed with the way its performing, shows a level of reasoning the previous models did not reach. I feel that it has some #Mythic elements to it as its vibecoded responses are impressive.
-
🧭 Esce Opus 4.8: un modello che non promette infallibilità, ma riconosce meglio i propri limiti. Un passo concreto verso un’AI più affidabile. #AI #Opus48
🔗 https://www.tomshw.it/business/esce-opus-48-un-modello-piu-onesto-quando-sbaglia
-
#Anthropic released #Opus48, the latest version of its advanced model, just 41 days after the previous version. The new model includes a feature called #DynamicWorkflows, designed to help larger models manage #complextasks. https://techcrunch.com/2026/05/28/anthropic-releases-opus-4-8-with-new-dynamic-workflow-tool/?eicker.news #tech #media #news
-
#Anthropic released #Opus48, the latest version of its advanced model, just 41 days after the previous version. The new model includes a feature called #DynamicWorkflows, designed to help larger models manage #complextasks. https://techcrunch.com/2026/05/28/anthropic-releases-opus-4-8-with-new-dynamic-workflow-tool/?eicker.news #tech #media #news
-
https://winbuzzer.com/2026/05/29/anthropic-ships-opus-48-with-dynamic-workflows-xcxwbn/
Anthropic has released Opus 4.8 and added Dynamic Workflows to Claude Code, bringing parallel subagents, resumable runs, and unchanged base pricing for teams.
#AI #Opus48 #DynamicWorkflows #ClaudeCode #Anthropic #Claude #GenAI #AIModels #AICoding #AIAgents #AgenticAI #EnterpriseAI
-
https://winbuzzer.com/2026/05/29/anthropic-ships-opus-48-with-dynamic-workflows-xcxwbn/
Anthropic has released Opus 4.8 and added Dynamic Workflows to Claude Code, bringing parallel subagents, resumable runs, and unchanged base pricing for teams.
#AI #Opus48 #DynamicWorkflows #ClaudeCode #Anthropic #Claude #GenAI #AIModels #AICoding #AIAgents #AgenticAI #EnterpriseAI
-
Tried out #AnthropicAI's #Opus48. Oddly, it frequently tells me (before making a decision) that it "wants [my] call on" things. That seems like a 2023-level idiom-mashup GPT mistake. "Make a call on..." and "It's your call" are all fine, but "WANT your call on"? That's weird.
-
Tried out #AnthropicAI's #Opus48. Oddly, it frequently tells me (before making a decision) that it "wants [my] call on" things. That seems like a 2023-level idiom-mashup GPT mistake. "Make a call on..." and "It's your call" are all fine, but "WANT your call on"? That's weird.