Die API-Preise der neuen Modelle GPT-5.5, Claude Opus 4.7 und Gemini 3.5 Flash verschleiern durch geänderte Tokenizer und interne Denkprozesse die wahren Betriebskosten.

GPT-5.5 verursacht effektive Kostensteigerungen von 49 bis 92 Prozent. Gemini 3.5 Flash erzeugt im Hintergrund permanente Thinking-Tokens, wodurch Benchmarks über 1.500 US-Dollar kosten.

#LLM #OpenWeights #GPT5 #GeminiFlash #AIGeneratedImage

https://www.all-ai.de/news/beitrage2026/kosten-ki-modell-real-1

Nuevo Gemini Omni, ¡La apuesta de Google por la multimodalidad completa!

00:00 Introducción
03:29 Probando Gemini Omni
12:25 Gemini 3.5 Flash
15:14 Otras novedades del Google I/O

Vía: Dotcsv - Divulgación sobre Inteligencia Artificial

#Divulgación #Ciencia #Tecnología #Ingeniería #Programación #RedesNeuronales #Robótica #IA #InteligenciaArtificial #DOTCSV #CarlosSantanaVega #GeminiOmni #GeminiFlash

https://www.youtube.com/watch?v=_WeP_t76xdo

qwant news | On-demand manufacturing startup SendCutSend raised $110M co-led by Sequoia, Paradigm, and Stripe co-founders Patrick and John Collison, valuing it at $1B

AI generated summary, Read the full article for complete information.

Google’s 2026 I/O unveiled a sweeping AI‑first overhaul, headlined by Gemini 3.5 Flash—the company’s fastest, most capable “agentic and coding” model, now powering the new AI‑Mode in Search and the Gemini app for long‑horizon workflows. Alongside Flash, Google introduced Gemini Omni for multimodal creation, Gemini Spark—a 24/7 personal AI agent that integrates with Gmail and Workspace, and the Universal Cart shopping hub built on a new Universal Commerce Protocol. Mobile updates include Android 17’s “Continue On” cross‑device handoff and the “Halo” status bar for AI agents. The company also refreshed its AI subscription tiers, lowering the top‑tier Ultra plan to $200/month and adding a $100‑per‑month entry tier, while expanding AI‑powered features across YouTube, Google Flow, and the new Android app ecosystem.

Read more: https://www.techmeme.com/260519/p68

#SendCutSend #Google #GeminiFlash #GeminiSpark #UniversalCart

RT @mweinbach: Wer sagt, dass TPUs nicht schnell sein können! Das ist in etwa Groq-Geschwindigkeit, aber aus einer TPU 8i, und bisher mit dem Gemini Flash-Modell deutlich intelligenter. Video Max Weinbach (@mweinbach) Google hat gerade eine Demo gezeigt: Das Gemini Flash-Modell läuft mit 600 bis 1400 Tokens pro Sekunde auf einer TPU 8i. Der Spitzenwert lag bei rund 1480 tok/s, im Durchschnitt bei etwa 800 tok/s — https://nitter.net/mweinbach/status/2056785716819145126#m

mehr auf Arint.info

#AIInnovation #GeminiFlash #GoogleAI #MachineLearning #TechDemo #TPU #arint_info

https://x.com/mweinbach/status/2056786997457625454#m

US Top News and Analysis | Google debuts new AI models, personal AI agents in effort to keep pace with OpenAI and Anthropic

AI generated summary, Read the full article for complete information.

Google unveiled a suite of new AI tools at its I O conference, aiming to stay competitive with OpenAI and Anthropic while expanding services for its massive user base. The centerpiece, Gemini 3.5 Flash, is a lighter‑weight, faster model that delivers cutting‑edge performance at roughly one‑third to one‑half the cost of comparable frontier models, and will become the default for the Gemini app and AI‑enabled search worldwide. A heavier Gemini 3.5 Pro is being tested internally and will roll out next month. Google also introduced Gemini Spark, a general‑purpose AI agent for the Gemini app that can reason across connected apps and act on user direction, currently in beta for trusted testers and AI Ultra subscribers. Additionally, the company revealed Omni, a world‑model that simulates physical environments and can edit video or generate realistic imagery within Flash, the Gemini app, Google Flow, and YouTube Shorts. These launches emphasize speed, lower latency, improved safety, and deeper product integration as Google seeks to broaden AI‑driven user experiences.

Read more: https://www.cnbc.com/2026/05/19/google-ai-ultra-gemini-spark-omni.html

#Google #OpenAI #SundarPichai #GeminiFlash #GeminiPro

Feed: All Latest | Google Search Goes Agentic—and Doesn’t Need You Anymore

AI generated summary, Read the full article for complete information.

Google unveiled a vision for a hyper‑personalized, agent‑driven Search that lets users create, customise and manage AI agents directly from the search interface, enabling tasks such as real‑time stock alerts, automated monitoring of sneaker releases, appointment booking and proactive information gathering even while users sleep; the rollout, tied to the new Gemini 3.5 Flash model and initially offered to AI Pro/Ultra subscribers, also introduces “vibe‑coded” results via the Antigravity tool, which can generate bespoke visual layouts, interactive mini‑apps and “super widgets” for complex planning, effectively shifting users from clicking through web links to delegating most of the browsing and execution work to Google‑hosted agents.

Read more: https://www.wired.com/story/google-search-goes-agentic-and-doesnt-need-you-anymore/

#Google #GeminiFlash #Duplex #I_Oconference #AIOverviews

구글 Nano Banana 3종 비교, NB2가 대부분 프로젝트에 최선인 이유

https://aisparkup.com/posts/10153

Search Live è ora disponibile in Italia tramite app Google o Google Lens. Sfrutta il nuovo modello Gemini Flash 3.1 Live per ricerche immediate e contestuali.

#searchlive #geminiflash
https://kiro.it/DvTA9

Một phiên bản mới của Nano Banana dựa trên mô hình Gemini Flash đang được thử nghiệm, hứa hẹn hiệu năng vượt trội hơn cả phiên bản tiền nhiệm Nano Banana Pro trước khi bị giới hạn. Nếu Google không "nerf" nó, người dùng có thể chưa sẵn sàng cho bước nhảy vọt này. AI tạo ảnh đang tiến bộ nhanh đến mức đáng lo ngại! #AI #GeminiFlash #NanoBanana #TríTuệNhânTạo #TạoẢnhAI

https://www.reddit.com/r/singularity/comments/1qstw03/upcoming_version_of_nano_banana_flash_outperforms/