Editorial archive

Releases & Models

News, tests and analysis to understand what changed, what is hype and what your business can already use.

Tela de radar escura com sinais luminosos
01 · Releases & Models

AI Radar 11/18/2026: Devin Bots arrive as leads of agent teams, Anthropic's IPO may price above range and Interact-0 learns to talk in groups

Cognition launches Devin Bots, agents that lead teams of agents, OpenClaw opens Sign in with ChatGPT, Anthropic may price its IPO above $80, Altman…

Ilustração editorial de planilhas e apresentações sobre mesa azul com folhas laranja e ampulheta, representando eficiência no trabalho.
02 · Releases & Models

CONFIRMED: GPT-6.2 Astra and Sol target professional work; Pro mode can raise costs

OpenAI launches Astra and Sol with document, coding and efficiency improvements. Astra has lower API rates, but OpenRouter’s Pro option can use more reasoning…

Tela de radar escura com sinais luminosos, ilustração do Radar de IA de 17/11/2026
03 · Releases & Models

AI Radar 11/17/2026: Anthropic prices its IPO tomorrow, ElevenLabs buys Japan's Kotoba and the Grok Bot team trades credits for feedback

5pm update: OpenAI launches GPT-6.2 Astra and Sol, Elon confirms a partial Grok 4.8 test on Grok Bot, SpaceXSI shows five Bots running its…

Tela de radar escura com sinais luminosos representando notícias de IA
04 · Releases & Models

AI Radar 11/16/2026: Claude Code's Code Mode rules the weekend, Cursor details the Fireworks breach and Zoom offers up to $150K for voice startups

5pm update: OpenAI opens applications on Nov 23 for free frontier model access for up to 200,000 scientists, Tencent's Hy4 flash is free on…

Ilustracao de uma rede tipo lattice de fisica de particulas representando o Modelo Padrao, evocando a colaboracao de pesquisa com o GPT-6 Astra
05 · Releases & Models

CONFIRMED: OpenAI uses GPT-6 Astra to help crack a 50-year-old physics problem

OpenAI launched a science blog detailing how its GPT-6 Astra model helped a University of California physicist propose a new mathematical definition for the…

Ilustração editorial de carteira violeta, cartão genérico e mão aprovando uma compra por impressão digital.
06 · Releases & Models

ChatGPT Wallet lets agents pay for purchases, but starts in the US

OpenAI’s wallet requires approval for each purchase and uses substitute card credentials. The rollout starts in dots for US Pro subscribers.

Tela de radar escura com sinais de voz e de chamadas de vídeo
07 · Releases & Models

AI Radar 11/13/2026: Claude joins Google Meet calls, Fish Agent ships a voice agent in 32 seconds and OpenAI turns on a limits reset

Claude Meetings started rolling out in beta on Team and Enterprise, joining Google Meet to write notes, docs and whiteboards live. Fish Audio launched…

Tela de radar escura com sinais luminosos de notícias de IA
08 · Releases & Models

AI Radar 12/11/2026: Gemini 4 Hydrogen officially launches, Grok Bot on Mac handles Messages and Mail, and Anthropic opens its IPO page

5pm update: Google launched Gemini 4 Hydrogen at $1.50/$7.50 per million tokens, with Orchestrator in Antigravity and already in Cursor; Grok Bot on Mac…

Prisma de vidro separa esferas coloridas em três caminhos ao lado de documentos, ilustração editorial de decisões e triagem com inteligência artificial.
09 · Releases & Models

CONFIRMED: Gemini Decision 1 makes AI triage cheaper

Google launches an API for structured decisions across text, images and audio at US$0.10 per million input tokens. Business uses and limits explained.

Tela de radar escura com pontos de luz conectados, capa do Radar de IA de 11/11/2026
10 · Releases & Models

AI Radar 11/11/2026: Grok Bot 1.0 leaves beta, Cursor agents coordinate across teams and Claude opens full apps in chat

5pm update: Grok Bot 1.0 is out, with shared bots in Slack, Teams and meetings, and Gopuff landed in Grok Bot. Fable 5.2 reached…

Esfera de vidro com rede de nós roxos e verdes, um caminho luminoso laranja e circuitos coloridos, ilustração editorial de inteligência artificial.
11 · Releases & Models

STRONG SIGNAL: ERNIE 5.5 Flash Preview appears on OpenRouter with million-token context

The catalog lists Baidu’s Preview with text API access and a large context window. Advertised pricing does not establish quality or a general release.

Loja em miniatura conectada a balões de conversa e vitrines digitais, ilustração editorial de empresas no ChatGPT.
12 · Releases & Models

CONFIRMED: OpenAI opens business listings in ChatGPT, but agents start in the US

Business Hub brings listings, product feeds and aggregated insights together. Business Agents initially serve eligible US companies.

Tela de radar escura com sinais luminosos
13 · Releases & Models

AI Radar 11/10/2026: Grok 4.8 heads to Grok Bot this week, Anthropic opens its IPO roadshow and OpenAI shows Control Room

Musk confirmed Grok 4.8 finished pre-training and gets a soft rollout to Grok Bot users this week. Anthropic launched its IPO roadshow and showed…

Radar escuro com anéis ciano e feixe de varredura
14 · Releases & Models

AI Radar 11/09/2026: Claude Fable 5.2 lands, Codex Pro with no 5-hour limit and Hermes Cloud in the browser

5pm update: Anthropic released Claude Fable 5.2 and Mythos 5.2 (already in Cursor), ChatGPT dots can now use several computers, Hermes Cloud opens in…

Radar de IA 06/11/2026: tela de radar escura com ondas de voz e rede de nós
15 · Releases & Models

AI Radar 11/06/2026: METR clocks Anthropic at 2x, OpenClaw swaps models without a release and LongCat free on Hermes

5pm update: METR estimates Anthropic is accelerating ~2x, OpenClaw discovers new models without a release, Nous makes LongCat-2.5 free for a week on the…

Ilustração editorial de um tênis coral em um miniestúdio, com uma fita luminosa de quadros de vídeo mostrando diferentes ângulos do produto.
16 · Releases & Models

Gemini Omni 1.2 Flash extends videos to 15 seconds and cuts pricing 30%

Google announces longer video generation and conversational editing at $0.07 per second in 720p. Real savings depend on how many attempts are needed.

Capa do Radar de IA 05/11/2026, estilo radar escuro
17 · Releases & Models

AI Radar 11/05/2026: Cat Wu at OpenAI, Meta buys Conductor and Claude Code cheapens subagents

Claude Code’s product lead joins OpenAI, Meta acquires Conductor (multi-agent coding orchestration), Claude Code cuts background-subagent status-line cost by about 99%, OpenAI shows a…

Ilustração editorial de smartphone e notebook exibindo criação de GIFs animados em um app de IA, com cores vibrantes em laranja, roxo e azul.
18 · Releases & Models

OpenAI launches Images 2.6, bringing animated GIFs into ChatGPT

OpenAI confirmed the global rollout of Images 2.6: ChatGPT and Codex gain animated GIF creation and saved reference images, while the API gets the…

Tela de radar escura com um ponto laranja brilhando sobre a América do Sul
19 · Releases & Models

AI Radar 11/04/2026: DevDay in São Paulo, Postman inside Claude Code and Codex and Tibo's ChatBLT

OpenAI DevDay Exchange hits São Paulo on Friday, Postman installed plugins into Claude Code and Codex without asking, a Codex told only to reboot…

Ilustracao editorial abstrata de uma esfera de rede neural vermelha e dourada abrindo um olho de camera, representando o modelo Hy4 da Tencent com visao computacional
20 · Releases & Models

Tencent launches Hy4 and says the open model enters the top tier of Chinese AI

Tencent launched Hy4, the new generation of its Hunyuan family, with first time computer vision and the same price as the previous version. The…

Tela de radar escura com pontos luminosos
21 · Releases & Models

AI Radar 11/03/2026: one allowance for ChatGPT and Codex, threads in the Claude API and your agents in meetings

ChatGPT and Codex will share one allowance from December, the Claude API gets message threads, Hermes v0.22.6 and OpenClaw 2026.10.10 ship, Tana brings your…

Ilustracao editorial abstrata de uma rede neural dourada sobre um wafer de chip, representando o modelo LongCat-2.5 da Meituan treinado em chips chineses
22 · Releases & Models

Meituan launches LongCat-2.5, a 1.6 trillion parameter model trained on Chinese chips

The Chinese delivery giant launched a 1.6 trillion parameter model trained entirely on domestic chips, with aggressive pricing and benchmarks placing it near Claude…

Tela de radar escura com feixe ciano e sinais luminosos
23 · Releases & Models

AI Radar 11/02/2026: Grok Bot hits its limits, Hermes imports your Grok Bot and Anthropic's numbers

Grok Bot limits tightened over costs, a Hermes command to migrate Grok Bot bots, Anthropic's S-1/A with third-quarter numbers, OpenAI dots getting more models…

Ilustracao abstrata digital representando um checkpoint de IA em teste numa arena publica de modelos, com fluxos de dados e balões de chat flutuantes em tons de azul e roxo
24 · Releases & Models

CONFIRMED: Google Launches Gemini 4 Carbon in General Availability

Google confirmed on Tuesday, November 10, 2026, that Gemini 4 Carbon, its most advanced model in cybersecurity and long agentic tasks, is now available…

Tela de radar escura com sinais brilhando, capa do Radar de IA de 30/10/2026
25 · Releases & Models

AI Radar 10/30/2026: Claude Ultra drops the 5-hour limit, SpaceXAI opens a startup program and a fake Claude Code extension hits Cursor

The day's roundup from X: Anthropic renames Max 20x to Ultra and drops the 5-hour limit for verified accounts, OpenAI announces Prism is shutting…

Ilustração editorial colorida representando edição de design assistida por inteligência artificial, com formas geométricas, camadas e elementos visuais flutuantes em estilo flat
26 · Releases & Models

Canva launches Brush, its own model that edits designs right inside the editor

CONFIRMED: Canva introduced Brush, its first vision language model trained to edit designs inside the editor itself, promising editing up to 69% faster and…

Tela de radar escura com varredura verde e pontos de sinal
27 · Releases & Models

AI Radar 10/29/2026: OpenAI doubles Pro limits, Phonic-1 tops voice and up to $500 in API credits

The day's roundup from X: OpenAI opens API credits for Pro and Business and, in the afternoon, reverses the cut and doubles Pro limits.…

Ilustração abstrata mostrando vários fluxos coloridos de dados de modelos de IA convergindo em um único ponto de luz, representando o GitHub Copilot Fusion orquestrando múltiplos modelos para um desenvolvedor
28 · Releases & Models

GitHub makes Copilot Fusion official and launches an app with Projects and Automations

CONFIRMED: at Universe 2026, GitHub took Copilot Fusion out of preview and launched a dedicated Copilot app alongside it, with multi repository Projects and…

Ilustração editorial de uma atendente com headset conversando com uma presença de IA representada por ondas sonoras luminosas, simbolizando um modelo de fala para fala em agente de voz
29 · Releases & Models

CONFIRMED: Rime launches Prism, a speech-to-speech model that kills the robotic pause in AI phone calls

Rime has launched Prism, a speech-to-speech model built for voice agents that cannot sound robotic: it listens to raw audio, replies in under a…

Tela de radar escura com sinais luminosos, capa do Radar de IA de 28/10/2026
30 · Releases & Models

AI Radar 10/28/2026: Gemini 4 Carbon takes the top spot, ChatGPT Pro cut and the Fable 5.5 that didn't ship

5 p.m. update: Google announced Gemini 4 Carbon, which debuts at the top of the Artificial Analysis index but for now only for cyber…

Mundo virtual interativo gerado por IA, com ondas de audio em paleta teal e ambar, simbolizando o Genie 4
31 · Releases & Models

CONFIRMED: Google DeepMind Launches Genie 4, a Real Time Interactive World With Native Audio

Google DeepMind officially announced Genie 4 on Tuesday, October 27, 2026, a world model that generates real time interactive environments from text or image…

Ilustração de radar com sinais de IA
32 · Releases & Models

AI Radar 10/27/2026: GPT-6.1 Terra, Anthropic's IPO, Gemini 4 Argon and the AGENTS.md warning

Today's roundup from X: OpenAI expands GPT-6.1 Terra, Anthropic files its IPO S-1, Cursor adds DeepSeek V4.1 Flash and an Agent SDK, Gemini 4…

Interface digital abstrata com núcleo luminoso e cursores convergindo, simbolizando um modelo de IA agindo em aplicativos
33 · Releases & Models

CONFIRMED: OpenAI launches GPT-6.1 Terra, the cheap model that uses your computer for you

OpenAI officially announced GPT-6.1 Terra, a fast and cheap model that understands what's on screen and acts across apps, now live in ChatGPT Work,…

Ilustração editorial de uma lua crescente feita de circuitos sobre um data center, com engrenagem 3D e gráfico financeiro holográficos, representando o Kimi K3.1 Pro da Moonshot AI
34 · Releases & Models

CONFIRMED: Moonshot launches Kimi K3.1 Pro, a 1M-context model that nears Claude Opus 5.5 on coding

Moonshot AI launched Kimi K3.1 Pro and a Highspeed version on Oct 23, already available in the app, Kimi Code and the API, with…

Ilustração editorial abstrata de uma rede neural saindo de um cofre digital em direção a mãos abertas, representando pesos de IA liberados ao público.
35 · Releases & Models

CONFIRMED: MiniMax releases M3.1 Flash, open-weight 428B parameter model with 1M token context

MiniMax publishes the full weights of M3.1 Flash on Hugging Face: a 428B parameter mixture-of-experts model with 23B active per token, 1M token context…

Ilustração editorial abstrata representando geração de vídeo por IA, com trilhas de movimento neon e quadros de filme se transformando em uma cena urbana futurista.
36 · Releases & Models

CONFIRMED: Kuaishou launches Kling 4.0, AI video model with Omni Reference for up to 15 elements

Kuaishou confirms the official launch of Kling 4.0 with Omni Reference, able to combine up to 15 references to keep character and scene consistent…

Ilustração abstrata de formas pixeladas se transformando em curvas vetoriais suaves, representando geração de SVG por IA
37 · Releases & Models

CONFIRMED: Ideogram launches Vector 1.0, its first editable SVG vector model

Ideogram confirmed the launch of Vector 1.0, its first model that generates editable SVG vector graphics from a prompt or converts existing images into…

Ilustração editorial de um agente de inteligência artificial trabalhando em segundo plano enquanto conecta laptop e celular
38 · Releases & Models

CONFIRMED: Anthropic launches Claude Duet, an agent that works alone and can shop for you

Anthropic has opened the preview of Claude Duet, a proactive agent that monitors your work, acts in the background and can even make business…

Ilustração editorial de torres de servidores de data center interligadas dentro de uma nuvem azul luminosa, representando os modelos da OpenAI entrando no Model Garden do Google Cloud
39 · Releases & Models

CONFIRMED: OpenAI brings Codex and its GPT-6 models to Google Cloud, the third hyperscaler to sell the brand

OpenAI and Google Cloud announced a partnership to bring GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna, plus Codex workflows, into Google's Model Garden. General…

Ilustração editorial de vozes conectadas a uma central de inteligência artificial, calendários e atendimento.
40 · Releases & Models

CONFIRMED: Voxtral 3 connects voice conversations to business tasks

Mistral combines transcription, speech and tool integration in a voice agent toolkit. Brazilian Portuguese presets are included, but full costs and reliability require testing.

Ilustração editorial de uma silhueta de inteligência artificial em rede neural brilhante, de perfil, com um balão de pensamento mostrando a silhueta de uma pessoa, representando a capacidade de intuir intenções não ditas do usuário
41 · Releases & Models

RUMOR: Anthropic is reportedly readying a model that reads users' unstated intentions better

The Information reports Anthropic is testing a frontier model built to infer what users actually mean, not just what they type, reviving speculation about…

Ilustração editorial abstrata de um painel financeiro digital representando o rumor de um recurso de finanças pessoais no Claude, da Anthropic, integrado ao Plaid
42 · Releases & Models

RUMOR: Claude may be preparing a personal finance feature with Plaid integration

Traces found inside the Claude app suggest Anthropic is testing a personal finance dashboard connected to Plaid, the service used to link bank accounts…

Ilustracao abstrata de tres feixes de luz coloridos representando os codinomes Aurora da OpenAI vazados e testados na Arena.
43 · Releases & Models

STRONG SIGNAL: leak points to GPT-Image-3, OpenAI's Aurora models caught generating images on Arena

Three model IDs named Aurora briefly appeared in OpenAI's API and were then spotted generating images on blind comparison platforms. Independent testers read this…

Ilustração editorial representando o Claude Dashboards e o Claude Motion da Anthropic, com gráficos de dados e linha do tempo de animação
44 · Releases & Models

Claude gets Dashboards and Motion: live data becomes a chart and a report becomes a video

Anthropic launched Claude Dashboards and Claude Motion in beta and took Docs, Slides and Design out of beta, making all three free to use…

Ilustração editorial de uma baleia digital luminosa representando a atualização do modelo DeepSeek V4.1 Flash 1017
45 · Releases & Models

CONFIRMED: DeepSeek updates V4.1 Flash and now beats Claude Opus 5 on coding and agent tasks

DeepSeek released V4.1 Flash 1017, a free open weight update that rises from 74.2% to 77% on the DeepSWE benchmark and now beats Claude…

Ilustração editorial representando o NVIDIA Nemotron 3.5 Super VL, modelo de IA multimodal aberto para agentes corporativos
46 · Releases & Models

CONFIRMED: NVIDIA open-weights Nemotron 3.5 Super VL for always-on multimodal agents

NVIDIA released Nemotron 3.5 Super VL, an open 120-billion-parameter model that reads text, image, document and video for always-on agents, with Scale AI, Mercor,…

Ilustração editorial de um cadeado aberto se transformando em rede neural brilhante, simbolizando a liberação de peso aberto de um modelo de IA, com silhueta de Mac Studio ao lado.
47 · Releases & Models

CONFIRMED: Meta open-weights Muse Spark, delivering the most capable US open model yet

Meta has open-weighted Muse Spark Open, an 816 billion parameter model with a 1 million token context window, built for coding and long-running agents…

Ilustração editorial do lançamento do FLUX 3 Image Klein, modelo de imagem de peso aberto da Black Forest Labs, com GPU e laptop estilizados
48 · Releases & Models

CONFIRMED: Black Forest Labs open sources FLUX 3 Image Klein, runs on consumer GPUs

Black Forest Labs officially launched FLUX 3 Image Klein, an open weights image model with 5.6 billion parameters that runs on a laptop with…

Ilustracao editorial representando onda de voz da IA Grok Voice 3.0 da xAI
49 · Releases & Models

CONFIRMED: xAI's Grok Voice 3.0 takes the lead in independent voice benchmark without a price hike

xAI officially launched Grok Voice 3.0 with a rare piece of independent validation: an Artificial Analysis benchmark puts the model at the top of…

Ilustracao editorial de uma esfera de rede neural translucida com poucos nos ativos acesos, flutuando sobre silhuetas de smartphone e notebook, representando agentes de IA rapidos e privados rodando no dispositivo
50 · Releases & Models

CONFIRMED: Liquid AI launches LFM2.5-24B-A2B, an open-weight model for fast on-device agents

Liquid AI released LFM2.5-24B-A2B, an open-weight mixture-of-experts model with only 2B active parameters, built for fast private agents running on-device.

Ilustracao editorial de um painel de catalogo de nuvem com identificadores de modelos de IA em destaque, representando o rastro do Claude Fable 5.5 encontrado no catalogo da Azure
51 · Releases & Models

STRONG SIGNAL: Claude Fable 5.5 surfaces in Azure's catalog before any Anthropic announcement

A line with the identifier claude-fable-5-5 appeared in Microsoft Azure's model catalog and in a LiteLLM configuration file, but Anthropic has not confirmed anything…

Ilustração editorial de uma estrela luminosa em degradê azul e violeta, no estilo do Gemini, irradiando luz sobre um globo terrestre conectado, simbolizando o lançamento global do Gemini 4 Argon pelo Google
52 · Releases & Models

CONFIRMED: Google opens Gemini 4 Argon to everyone and rejoins the frontier race

After delays and turmoil at DeepMind, Argon leaves restricted access and reaches the API, app and Search, priced at US$2 and US$10 per million…

Ilustração editorial de uma estrutura ciano conectando documentos, calendário e código, representando memória e foco de agentes de IA.
53 · Releases & Models

CONFIRMED: Meta Muse Wonder bets on agents that remember and persist

Meta launches Muse Wonder and Wonder Full with memory across conversations and a focus on completing tasks. Access starts through the API and Muse…

Ilustração editorial sobre a expansão da janela de contexto de um modelo de IA para 500 mil tokens
54 · Releases & Models

CONFIRMED: OpenAI nearly doubles GPT-6.1 Sol's default context in Codex and ChatGPT Work

An OpenAI executive confirmed on X that GPT-6.1 Sol now defaults to 500,000 tokens of context in Codex and ChatGPT Work, nearly 80% more…

Ilustracao editorial sobre o rumor do modelo Claude Next dentro do Claude Code
55 · Releases & Models

RUMOR: Testers say they spotted an unannounced Claude inside Claude Code, calling it "Claude Next"

A tester with a track record of spotting unreleased models says he found, inside Claude Code, a model that makes Opus 5.5 look primitive.…

Ilustração editorial abstrata representando um agente de IA único conectado a documentos, e-mail, código e chat, em paleta de cores do Google com toque laranja da Anthropic
56 · Releases & Models

CONFIRMED: Google launches a single Gemini agent, already running on Anthropic's Claude

At the Gemini at Work 2026 event, Google Cloud unveiled a single agent that takes objectives and returns finished work across Workspace, Microsoft 365…

Ilustração editorial de rotas luminosas conectando documentos e um terminal a um núcleo de inteligência, representando o acesso ao Step 5 Preview pelo OpenRouter.
57 · Releases & Models

STRONG SIGNAL: Step 5 Preview appears on OpenRouter for agents

The catalog lists a 1-million-token context and output at $2.70 per million. The change is an access route, not a new model launch or…

Ilustração editorial: resposta de chatbot se transforma em gráfico, mapa, botões e calculadora interativos em celular e notebook
58 · Releases & Models

CONFIRMED: GPT-6 reaches free ChatGPT as answers turn into charts, maps and tools

OpenAI brought GPT-6 to every ChatGPT plan: Sol for subscribers since October 7 and Luna for Free and Go from October 8. The new…

Ilustração editorial de uma banana dourada e cartões com um frasco de cosmético fictício em cenários coloridos, simbolizando edição de imagens de produto.
59 · Releases & Models

CONFIRMED: Google Nano Banana 2.1 cuts campaign image costs, but requires an October migration

Google released Nano Banana 2.1 on October 6, with more consistent editing and a lower 1K image-output charge. Its predecessor enters the retirement schedule…

Ilustração editorial de pássaro de origami coral organizando cartões azuis em um pequeno centro de triagem, metáfora da automação eficiente com modelos de IA menores.
60 · Releases & Models

CONFIRMED: Claude Haiku 5.5 lowers automation costs, but long prompts change the bill

The new Haiku adds adjustable reasoning, a 1 million token context and lower prices for short tasks. Anthropic also cuts Sonnet cache read pricing…

Ilustração editorial de reunião com quatro pessoas; ondas de áudio coloridas saem de cada participante para uma transcrição com falas identificadas por falante (S1 a S4).
61 · Releases & Models

Gemma 4 based model targets cleaner transcripts for multi-speaker meetings

STRONG SIGNAL: google/DiarizationLM-Gemma-4-E4B-v1 appeared on Hugging Face as a Gemma 4 E4B fine-tune for correcting turn boundaries and speaker labels in transcripts. The page…

Ilustração editorial do lançamento do Kolibri, modelo de IA aberto da Aleph Alpha, com referências visuais à bandeira alemã e à União Europeia combinadas a uma rede neural abstrata
62 · Releases & Models

CONFIRMED: Aleph Alpha launches Kolibri, giving Europe its first open model that truly competes

Aleph Alpha, the German lab based in Heidelberg, released Kolibri on October 3, the anniversary of German reunification: an open weight model under the…

Ilustração editorial sobre inteligência artificial segura para uso governamental
63 · Releases & Models

CONFIRMED: Claude for Government exits beta and reaches general availability under the US government's toughest security bar

Anthropic ended the public beta that opened in July and made Claude for Government permanently available to US federal and state agencies, running under…

Ilustração editorial mostrando a transição dos Gems do Google Gemini para o novo recurso Skills, com a estrela do Gemini ao centro conectando os dois ícones
64 · Releases & Models

CONFIRMED: Google retires Gemini Gems, migrating everything to Skills in November

Automatic migration starts November 17, but Skills still can't replicate Gems' video, music, Canvas and Deep Research, and hasn't reached Europe yet.

Arte editorial abstrata sobre velocidade e integração de agentes no lançamento do Claude Sonnet 5.5
65 · Releases & Models

CONFIRMED: Claude Sonnet 5.5 speeds up agents but demands migration care

CONFIRMED: Anthropic launched Claude Sonnet 5.5 with more than 30% faster output, the same price as Sonnet 5, and API changes that can break…

Ilustração editorial sobre o vazamento de um modelo intermediário e secreto de inteligência artificial
66 · Releases & Models

STRONG SIGNAL: OpenAI leaks Astra Minor, a fourth model in the GPT-6 family

The same day it shipped GPT-6 Sol and Luna, leaked Help Center documentation and cloud configuration point to a fourth model, Astra Minor, positioned…

Ilustração editorial sobre uma plataforma de hardware para conter agentes de inteligência artificial fora de controle
67 · Releases & Models

CONFIRMED: Nvidia launches platform to contain runaway AI agents, two days after OpenAI's containment failure

Nvidia launched the Open Agent Safety Platform, a dual layer system that promises to contain runaway AI agents directly at the hardware level. The…

Ilustracao editorial do assistente de IA sempre ativo rumorado da OpenAI, codinome 'o', mostrando um motivo de orbe luminoso com fluxos de dados continuos em fundo degrade azul escuro e laranja.
68 · Releases & Models

CONFIRMED: OpenAI launches Dots, its always-on agent to rival Meta's Muse

At its September 29 DevDay, OpenAI confirmed the name of the always-on assistant that last week's leaks had called "o": the official product is…

Ilustração editorial mostrando o Gemini 3.8 Live com avatar animado sincronizado por voz do Google
69 · Releases & Models

CONFIRMED: Google launches Gemini 3.8 Live with a talking avatar for enterprises

On September 24, Google launched Gemini 3.8 Live with Live Avatar, pairing an animated, voice-synced face with its real-time dialogue model, available to Gemini…

Ilustração editorial do ChatGPT Pro 500 com velocidade Ultrafast da OpenAI em chips Cerebras
70 · Releases & Models

CONFIRMED: OpenAI launches ChatGPT Pro 500 with speed powered by Cerebras chips

OpenAI confirmed at DevDay the ChatGPT Pro 500, a $500 monthly subscription with exclusive access to Ultrafast, a mode that speeds up GPT-6 Astra…

Logo oficial da OpenAI sobre foto de cibersegurança, capa da notícia: SINAL FORTE: OpenAI prepara o GPT-6 Cyber, quarto modelo de ciberdefesa do ano, para o DevDay
71 · Releases & Models

CONFIRMED: OpenAI reveals official GPT-6 Cyber date and launches Daybreak Red for critical infrastructure

Co-founder Greg Brockman confirmed on X that November 12 is GPT-6 Cyber day, the cybersecurity model OpenAI already uses to help defend projects like…

Logo oficial da OpenAI sobre foto de inteligência artificial, capa da notícia: RUMOR: OpenAI prepara segundo modelo com arquitetura em loop para o DevDay de 29 de setembro
72 · Releases & Models

RUMOR: OpenAI reportedly prepping a second looped-architecture model for its September 29 DevDay

A single X post claims OpenAI will ship a second recurrent-depth (looped) model at its September 29 DevDay, the same technical family associated with…

Marketplace da Anthropic para Claude
73 · Releases & Models

CONFIRMED: Anthropic launches Claude Marketplace with more than 2,000 connectors

CONFIRMED: Anthropic launched Claude Marketplace, bringing together more than 2,000 connectors, partner products and consulting firms in a single catalog. The move simplifies buying…

CONFIRMADO: Meta amplia os óculos Ray-Ban com áudio, Gen3 e conexão com o Muse
74 · Releases & Models

CONFIRMED: Meta expands Ray-Ban glasses with audio, Gen3 and Muse connection

Meta announced new Ray-Ban Meta Audio and Ray-Ban Meta Gen3 glasses, focusing on audio, battery life, camera and future integration with the Muse personal…

Alibaba Zhenwu V900 e plano Qwen
75 · Releases & Models

CONFIRMED: Alibaba launches Zhenwu V900 chip, aims for a 10 trillion parameter Qwen4

CONFIRMED: at the September 22 Apsara Conference, Alibaba announced the Zhenwu V900 AI chip, with triple the performance of its predecessor and support for…

Intrinsic Core para robótica industrial
76 · Releases & Models

CONFIRMED: Google's Intrinsic open sources the core of its industrial robotics platform

CONFIRMED: Intrinsic, Alphabet's robotics unit, released Intrinsic Core on September 22 under an Apache 2.0 license, with real time control, part perception, motion and…

CONFIRMADO: Anthropic lança Claude Opus 5.5 com mais capacidade e custo menor
77 · Releases & Models

CONFIRMED: Anthropic launches Claude Opus 5.5 with greater capability at lower cost

CONFIRMED: beyond greater capability and 40% lower cost on typical workloads, Claude Opus 5.5 documentation reveals integration changes that may break existing agents, especially…

CONFIRMADO: Mistral Small 4 chega ao Firefox Smart Window e leva escolha de modelos ao navegador
78 · Releases & Models

CONFIRMED: Mistral Small 4 comes to Firefox Smart Window, bringing model choice to the browser

Mozilla and Mistral confirmed a partnership bringing Mistral Small 4 to Firefox Smart Window, expanding the beta to France and making model choice a…

RUMOR: vazamento aponta vozes personalizadas e novo modelo Avocado para o Muse, da Meta
79 · Releases & Models

RUMOR: leak points to custom voices and new Avocado model for Meta's Muse

A report from tracker TestingCatalog, published on September 10, describes internal tests of custom voices, a voice mode and reasoning controls inside Meta's personal…

SINAL FORTE: código vazado mostra OpenAI construindo Managed Agents antes da DevDay
80 · Releases & Models

STRONG SIGNAL: leaked code shows OpenAI building Managed Agents ahead of DevDay

TestingCatalog found unfinished screens inside OpenAI's own platform for a product called Managed Agents, which would host both the agent and its execution environment…

RUMOR: vazamento aponta que a OpenAI já teria treinado o “Bel”, pré-treino gigante sucessor do “Doug”
81 · Releases & Models

RUMOR: leak claims OpenAI already trained "Bel", a giant pretraining run succeeding "Doug"

A single X account claims OpenAI has finished pretraining a model called Bel, with over 10 trillion parameters, the successor to the Doug pretraining…

CONFIRMADO: Kimi K3, modelo aberto de 2,8T da Moonshot AI, chega ao Amazon Bedrock em disponibilidade geral
82 · Releases & Models

CONFIRMED: Kimi K3, Moonshot AI's 2.8T open model, goes GA on Amazon Bedrock

AWS confirmed on September 18, 2026 that Moonshot AI's 2.8 trillion parameter open weight Kimi K3 is generally available on Amazon Bedrock in every…

CONFIRMADO: Alibaba lança Qwen3.8-Omni-Flash, primeiro modelo omnimodal agentic da Qwen a preço de Flash
83 · Releases & Models

CONFIRMED: Alibaba launches Qwen3.8-Omni-Flash, Qwen's first agentic omni-modal model at Flash pricing

On September 18, 2026, Alibaba launched Qwen3.8-Omni-Flash, the first Qwen model to combine text, image, audio and video with full agentic capability, pricing it…

GLM-5.3-FlashX: Z.ai lança modelo de código a 200 tokens por segundo
84 · Releases & Models

GLM-5.3-FlashX: Z.ai launches a coding model at 200 tokens per second

STRONG SIGNAL: Z.ai launched GLM-5.3-FlashX, a faster paid variant of GLM-5.3-Flash reaching up to 200 tokens/s, documented officially and confirmed across OpenRouter and independent…

Claude Code 2.1.278 acaba com a cobrança oculta do classificador de modo automático
85 · Releases & Models

Claude Code 2.1.278 ends the hidden charge for the auto mode classifier

CONFIRMED: Anthropic released Claude Code 2.1.278, moving auto mode's safety classifier to run server-side at no extra charge, with a warning when a session…

CONFIRMADO: Google transforma o CC em agente compartilhado para famílias
86 · Releases & Models

CONFIRMED: Google turns CC into a shared AI agent for families

Google expanded CC, its experimental agent, to families and groups of up to six people. The update gives the agent its own Google Account,…

Claude Code 2.1.277 amplia instruções de projeto e endurece agentes
87 · Releases & Models

Claude Code 2.1.283 strengthens governance, reliability, and project instructions

CONFIRMED: Claude Code 2.1.283 expands control over models, telemetry, plugins, and gateways, fixes SDK, MCP, fallback, and cloud-session failures, and preserves useful 2.1.277 history.

RUMOR: Slug claude-opus-5-2 aparece no Microsoft Foundry e acende aposta em novo Opus da Anthropic
88 · Releases & Models

RUMOR: claude-opus-5-2 slug surfaces on Microsoft Foundry, fueling bets on a new Anthropic Opus

An identifier reading claude-opus-5-2 reportedly surfaced in the Microsoft AI Foundry catalog, and a single X account describes silent routing inside Claude Code. Anthropic…

Google libera Gemini 3.8 Live e coloca raciocínio em segundo plano na voz
89 · Releases & Models

CONFIRMED: Google launches Gemini 3.8 Live with background reasoning for voice

Google launched Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, voice models that run tools and reason in the background without interrupting conversation.…

CONFIRMADO: Musk diz que o Grok 4.8 termina o treino e entra em RL nesta semana
90 · Releases & Models

CONFIRMED: Musk says Grok 4.8 will finish training and enter RL this week

Elon Musk said Grok 4.8 will finish pretraining and enter reinforcement learning this week, but xAI has not announced a public endpoint or release…

CONFIRMADO: GPT-5.5 sai do ChatGPT em 14 de outubro, mas continua na API
91 · Releases & Models

CONFIRMED: GPT-5.5 leaves ChatGPT on October 14 but stays on the API

OpenAI confirmed that GPT-5.5 will be removed from ChatGPT, ChatGPT Work, and plan-authenticated Codex on October 14, while remaining available through the API and…

Claude Code 2.1.273 melhora gateways, MCP e sessões remotas
92 · Releases & Models

Claude Code 2.1.273 improves gateways, MCP, and remote sessions

CONFIRMED: Anthropic released Claude Code 2.1.273 with improvements to gateways, MCP reconnection, remote control, permissions, and agent-result delivery.

Ilustração editorial de dois cristais luminosos conectados, um protegido por cúpula e outro em câmara científica, representando Fable e Mythos com salvaguardas distintas.
93 · Releases & Models

CONFIRMED: Claude Fable 5.2 cuts prices; Mythos access stays restricted

Anthropic confirms Fable 5.2 and Mythos 5.2 at $8 for input and $40 for output per million tokens, and opens a launch offer with…

Claude Code 2.1.270 corrige regressão que pedia permissão para comandos Git seguros
94 · Releases & Models

Claude Code 2.1.270 fixes regression that asked permission for safe Git commands

CONFIRMED: Claude Code 2.1.270 fixes a 2.1.269 regression that made read-only Git commands in Bash ask for permission after a long session.

OpenAI revela como o Habitat passou a sustentar 1 bilhão de usuários do ChatGPT
95 · Releases & Models

OpenAI reveals how Habitat grew to support 1 billion ChatGPT users

CONFIRMED: OpenAI described how it evolved Habitat from a Python library into a globally distributed storage platform serving 1 billion ChatGPT users and 22…

OpenAI diz ter resolvido Navier Stokes com sistema de agentes além do GPT 6 Astra
96 · Releases & Models

Navier Stokes turns into a dispute: mathematician accuses OpenAI of using his work and threatening his career

The claim that OpenAI solved the Navier Stokes Millennium Prize problem turned into a public dispute after mathematician Tristan Buckmaster accused the company of…

Claude Code 2.1.268 mira o custo dos gateways e fecha brechas em agentes
97 · Releases & Models

Claude Code 2.1.268 targets gateway costs and closes agent security gaps

CONFIRMED: Claude Code 2.1.268 adds price synchronization for managed gateways, strengthens security controls, and fixes failures affecting compatible endpoints, sessions, MCP, and CPU usage.

CONFIRMADO: DeepSeek lança V4.1 Flash com visão nativa e derruba o preço da API
98 · Releases & Models

CONFIRMED: DeepSeek launches V4.1 Flash with native vision and cuts API prices

DeepSeek has confirmed the launch of V4.1 Flash, the smallest model in its new architecture family, with native visual understanding, a 1 million token…

SINAL FORTE: Alexandr Wang confirma o codinome Watermelon, o próximo modelo de fronteira da Meta
99 · Releases & Models

STRONG SIGNAL: Alexandr Wang confirms Watermelon, Meta's next frontier model codename

Days after shipping Muse Spark 1.3 and claiming parity with Fable 5.1 and Opus 5 on independent benchmarks, Meta's Chief AI Officer Alexandr Wang…

Cristal violeta conectado a documentos, imagem e vídeo diante de uma porta aberta, ilustração sobre acesso a inteligência artificial multimodal.
100 · Releases & Models

STRONG SIGNAL: Qwen4 Max appears on OpenRouter with usage pricing

OpenRouter lists Qwen4 Max with token-based pricing. The new route is a strong signal, while official documentation retains specific preview conditions.

Meta apresenta o Muse, agente pessoal que quer agir por você
101 · Releases & Models

Meta introduces Muse, a personal agent built to act on your behalf

Meta introduced Muse, a personal agent that browses, fills out forms and keeps working after users close the app. The rollout starts in the…

GPT Image 2.5 ganha suporte oficial no SDK Python da OpenAI
102 · Releases & Models

GPT Image 2.5 gets official support in OpenAI’s Python SDK

CONFIRMED: version 3.10.0 of OpenAI’s Python SDK added GPT Image 2.5 models and image options, turning the visual launch into an official integration surface…

RUMOR: DeepSeek V5 seria construído do zero para rivalizar com o Claude Mythos
103 · Releases & Models

RUMOR: DeepSeek V5 reportedly built from scratch to rival Claude Mythos

A post on X from leak tracker Mr_Salio, circulated on September 7, 2026, claims DeepSeek is building V5 from scratch, targeting Anthropic's Mythos tier…

iFlytek lança o Spark X2.5, modelo de 293B treinado do início ao fim em chip chinês
104 · Releases & Models

iFlytek launches Spark X2.5, a 293B model trained end to end on Chinese chips

iFlytek officially launched Spark X2.5 this Monday, a 293 billion parameter MoE model focused on coding and agents, trained and served entirely on domestic…

Logo da OpenAI sobre fundo gradiente azul-roxo, capa da noticia do GPT-5.6
105 · Releases & Models

CONFIRMED: GPT-6.1 Sol targets Astra-level work at one-fifth the price

OpenAI announced GPT-6.1 Sol, an upgrade that brings the model closer to GPT-6 Astra on coding, computer use and professional work, with API prices…

Claude prova o Último Teorema de Fermat em Lean, com verificação automática
106 · Releases & Models

Claude proves Fermat's Last Theorem in Lean, with automatic verification

Anthropic confirmed that its Claude model, orchestrated through the Prove2Me platform, produced the first complete, machine verified proof of Fermat's Last Theorem in the…

RUMOR: strings vazadas “Marshmallow” e “Melon” insinuam Claude Opus 5.1 e Sonnet 5.1
107 · Releases & Models

RUMOR: leaked strings "Marshmallow" and "Melon" hint at Claude Opus 5.1 and Sonnet 5.1

Two early-access identifiers, claude-marshmallow-eap and claude-melon-eap, surfaced in developer tools in late August, in the same naming pattern that preceded Claude Opus 5. Community…

Logo oficial da Anthropic sobre foto de inteligência artificial, capa da notícia: CONFIRMADO: Anthropic lança Claude Fable 5.1 e Claude Mythos 5.1 com preço menor e novos limites
108 · Releases & Models

CONFIRMED: Anthropic launches Claude Fable 5.1 and Claude Mythos 5.1 with lower cost and new limits

Anthropic launched Claude Fable 5.1 for general availability and Claude Mythos 5.1 for trusted access programs. The same model combines cheaper cache reads, adjusted…

Capa editorial Max IA Assistant com o tema K2 Horizon sobre foto de inteligência artificial, notícia: Institute of Foundation Models lança K2 Horizon, a maior leva de modelos de IA totalmente abertos
109 · Releases & Models

Institute of Foundation Models Launches K2 Horizon, the Largest Fully Open AI Model Release Yet

The Institute of Foundation Models, affiliated with MBZUAI, released K2 Horizon on September 3, 2026, a family of six AI models ranging from 0.9…

Logo oficial da Microsoft sobre foto de voz e áudio, capa da notícia: Microsoft lança VibeVoice-ASR-Streaming para transcrever e identificar falantes ao vivo
110 · Releases & Models

Microsoft launches VibeVoice-ASR-Streaming to transcribe and identify speakers live

CONFIRMED: Microsoft launches VibeVoice-ASR-Streaming, an open model combining live transcription and speaker attribution in ten languages, including Portuguese.

Logo oficial da Meta sobre foto de código e desenvolvimento, capa da notícia: Meta lança o Muse Spark 1.3 e vende o salto em código e agente
111 · Releases & Models

Meta launches Muse Spark 1.3 and sells a jump in coding and agents

Meta put Muse Spark 1.3 in Muse Code and the API at 1.2 pricing. Zuckerberg calls it the house's biggest jump. Max reasoning has…

Capa editorial Max IA Assistant com o tema Multiverse · Quasar 438B sobre foto de inteligência artificial, notícia: Multiverse lança o Quasar 438B e vende o título de líder europeu
112 · Releases & Models

Multiverse launches Quasar 438B and sells a European leader title

Spain's Multiverse Computing published Quasar 438B, a 438 billion parameter model for agents and coding. Artificial Analysis scores it 43, above Mistral Medium 3.5.…

Logo oficial da Gemini sobre foto de inteligência artificial, capa da notícia: Gemini liga leitura agentic de vídeo
113 · Releases & Models

Gemini turns on agentic video understanding

Google turned on agentic video understanding in Gemini 3.7, 3.6 Flash and 3.5 Flash-Lite. Vendor numbers: up to 88% fewer tokens and 66% lower…

Logo oficial da xAI sobre foto de publicidade e preços, capa da notícia: xAI lança Grok 4.7 com foco em tarefas longas e preço de US$ 2 por milhão de tokens
114 · Releases & Models

CONFIRMED: xAI launches Grok 4.7 for long-running tasks at $2 per million input tokens

xAI announced Grok 4.7 on September 21, with a larger base model, longer reinforcement training, and a focus on coding and knowledge work. Independent…

Logo oficial das marcas Qwen e OpenRouter sobre foto de inteligência artificial, capa da notícia: Qwen3.8-Max-0902 ganha confirmação independente e chega à OpenRouter
115 · Releases & Models

Qwen3.8-Max-0902 gets independent confirmation and reaches OpenRouter

CONFIRMED: Qwen3.8-Max-0902 keeps the same pricing and context window, received independent validation on Code Arena WebDev, and is scheduled to become the Qwen3.8 Max…

Logo oficial da DeepSeek sobre foto de inteligência artificial, capa da notícia: DeepSeek abre o V4-Flash-Vision-Exp, o primeiro olho da família V4
116 · Releases & Models

DeepSeek open-sources V4-Flash-Vision-Exp, the V4 family's first eyes

DeepSeek published DeepSeek-V4-Flash-Vision-Exp on Hugging Face, the first experimental multimodal model in the V4 family, at 305B parameters under an MIT license. The pitch…

Logo oficial da OpenAI sobre foto de saúde, capa da notícia: ChatGPT passa a conectar prontuário Epic e bases oficiais de saúde
117 · Releases & Models

ChatGPT now connects Epic records and official healthcare sources

On September 1, 2026 OpenAI launched an Epic EHR integration and a Healthcare Public Data plugin inside ChatGPT for Healthcare, with read-only access to…

Logo oficial da Gemini sobre foto de cibersegurança, capa da notícia: Google lança o Gemini 3.8 Flash e reserva o Cyber para defesa
118 · Releases & Models

Google launches Gemini 3.8 Flash and keeps Cyber for defenders

Google confirmed Gemini 3.8 Flash at 3.7 introductory pricing and gated 3.8 Flash Cyber behind Fairwind. This morning's ID is now a product.

Logo oficial da Meta sobre foto de voz e áudio, capa da notícia: Meta lança Muse Voice Transcribe, o primeiro ouvido em tempo real da Superintelligence Labs
119 · Releases & Models

Meta launches Muse Voice Transcribe, Superintelligence Labs' first real-time ear

Meta Superintelligence Labs launched Muse Voice Transcribe on September 1, 2026, its first real-time audio perception model: streaming ASR, diarization for 20-plus speakers and…

Logo oficial da Claude sobre foto de código e desenvolvimento, capa da notícia: Claude Fable 5.1 atualiza agentes longos, ciência e economia do contexto
120 · Releases & Models

Claude Fable 5.1 updates long-running agents, science and context economics

Anthropic updated Fable 5.1 and Mythos 5.1 with 75% cheaper cache reads, new agent controls and science advances, but migration breaks integrations and retention…

Logo oficial das marcas Nvidia e xAI sobre foto de espaço, capa da notícia: Nvidia lança o primeiro CPU feito para agentes de IA, e a SpaceXAI já vai usá-lo até em órbita
121 · Releases & Models

Nvidia Launches the First CPU Built for AI Agents, and SpaceXAI Will Use It Even in Orbit

Nvidia announced on August 24, 2026 that SpaceXAI will deploy Vera, its first CPU designed specifically for artificial intelligence agents, to accelerate its next…

Logo oficial da Mistral AI sobre foto de data center e infraestrutura, capa da notícia: Mistral monta coalizão de empresas europeias para bancar até 1 gigawatt de computação soberana até 2030
122 · Releases & Models

Mistral Assembles a Coalition of European Companies to Fund Up to 1 Gigawatt of Sovereign Compute by 2030

Mistral AI announced three moves aimed at European digital sovereignty on August 11, 2026: regional inference endpoints letting customers choose whether their data is…

Logo oficial da Nvidia sobre foto de mercado financeiro, capa da notícia: Nvidia paga US$ 6 bilhões à Poolside para construir um modelo de IA aberto rival da OpenAI e da Anthropic
123 · Releases & Models

Nvidia Pays $6 Billion to Poolside to Build an Open Weight AI Model Rivaling OpenAI and Anthropic

Nvidia has struck a $6 billion licensing deal with startup Poolside to use its model building technology, called Model Factory, while also investing another…

Logo oficial das marcas AWS e OpenAI sobre foto de cibersegurança, capa da notícia: AWS amplia acesso aos modelos GPT-5.6 da OpenAI com inferência entre regiões e modelos de cibersegurança
124 · Releases & Models

AWS Expands Access to OpenAI's GPT-5.6 Models With Cross-Region Inference and Cybersecurity Models

AWS began offering cross-region inference for OpenAI's GPT-5.6 Sol, Terra and Luna models within Amazon Bedrock on August 17, 2026, according to an official…

Logo oficial da OpenAI sobre foto de inteligência artificial, capa da notícia: OpenAI lança ChatGPT for Teens com proteções ativadas por padrão e controle parental
125 · Releases & Models

OpenAI Launches ChatGPT for Teens With Default Protections and Parental Controls

OpenAI began rolling out ChatGPT for Teens globally on August 18, 2026, a version of the product built for users ages 13 to 17…

Logo oficial da Cerebras sobre foto de chips e semicondutores, capa da notícia: Cerebras lança CS-4, sistema que promete até 30 vezes mais velocidade de inferência que GPUs
126 · Releases & Models

Cerebras CS-4: what the up-to-30x faster inference claim means

Cerebras announced CS-4 on August 18, 2026. The system combines three WSE-3 Turbo processors on the Nexus platform. The company reports over 4,400 tokens…

Logo oficial das marcas xAI e GitHub Copilot sobre foto de código e desenvolvimento, capa da notícia: xAI leva o Grok 4.6 ao GitHub Copilot em oito ambientes de desenvolvimento diferentes
127 · Releases & Models

xAI Brings Grok 4.6 to GitHub Copilot Across Eight Development Surfaces

Two days after launching Grok 4.6 on August 12, 2026, xAI announced on August 14 that the model had arrived in GitHub Copilot, selectable…

Logo oficial da Gemini sobre foto de código e desenvolvimento, capa da notícia: Google lança Gemini 3.7 Flash, modelo barato para código e agentes de IA
128 · Releases & Models

Google Launches Gemini 3.7 Flash, a Cheap Model for Code and Agents

Google launched Gemini 3.7 Flash on August 13, 2026, a workhorse model built for software engineering, AI agents and multi-step task execution. According to…

Logo oficial da Gemini sobre foto de apps e dispositivos, capa da notícia: Google lança Pixel 11 com chip Tensor G6 e Gemini capaz de agir em mais de 40 aplicativos
129 · Releases & Models

Google Launches Pixel 11 With Tensor G6 Chip and Gemini Acting Across 40+ Apps

Google unveiled the Pixel 11 lineup on August 12, 2026, featuring the 2 nanometer Tensor G6 chip and Gemini Intelligence carrying out multistep tasks…

Logo oficial da DeepSeek sobre foto de código e desenvolvimento, capa da notícia: DeepSeek tira o V4 Pro da fase de testes e mira agentes de IA
130 · Releases & Models

DeepSeek Moves V4 Pro Out of Preview, Targets AI Agents

DeepSeek launched the production version of V4 Pro, dubbed 0813, on August 12, 2026, ending nearly four months in preview. According to Unite.AI and…

Logo oficial das marcas Microsoft e Nvidia sobre foto de chips e semicondutores, capa da notícia: Microsoft prepara lançamento do chip de IA Maia 300 para o outono, mirando menos dependência da Nvidia
131 · Releases & Models

Microsoft Preps Maia 300 AI Chip Reveal for the Fall, Aiming to Cut Nvidia Dependence

Microsoft plans to publicly unveil the Maia 300, its new AI accelerator chip, as soon as this fall, possibly in September 2026, according to…

Logo oficial da Meta sobre foto de inteligência artificial, capa da notícia: Zuckerberg publica manifesto sobre superinteligência e Meta lança Muse Glimmer em código aberto
132 · Releases & Models

Zuckerberg Publishes Superintelligence Manifesto as Meta Launches Open Source Muse Glimmer

On August 10, 2026, Mark Zuckerberg published the essay 'The Future is for Everyone,' roughly 6,500 words long, arguing that superintelligence should not be…

Logo oficial da xAI sobre foto de inteligência artificial, capa da notícia: xAI aposenta grok-imagine-image-quality e migra chamadas para o Image 2.0
133 · Releases & Models

xAI Retires grok-imagine-image-quality and Moves Requests to Image 2.0

CONFIRMED: xAI will retire the grok-imagine-image-quality identifier on November 2, 2026. Requests will keep responding, but they will be served by grok-imagine-image-2.0 at low…

Logo oficial da Meta sobre foto de código e desenvolvimento, capa da notícia: Meta lança Muse Code, seu primeiro agente de codificação com IA
134 · Releases & Models

Meta Launches Muse Code, Its First AI Coding Agent

Meta launched Muse Code, its first AI coding agent, in public beta on August 5, 2026, for macOS and Linux. Built on the new…

Logo oficial da xAI sobre foto de inteligência artificial, capa da notícia: Grok 4.6: xAI lança novo modelo e já prepara o Grok 4.7
135 · Releases & Models

Grok 4.6: xAI Launches New Model, Already Preps Grok 4.7

xAI released Grok 4.6 on August 7, 2026, hitting the deadline Elon Musk had set. The new model keeps Grok 4.5's 1.5 trillion parameter…

Ilustração editorial de pesquisadores colaborando com computadores e instrumentos científicos em uma biblioteca, sob um globo luminoso conectado.
136 · Releases & Models

CONFIRMED: OpenAI doubles free ChatGPT research target, but switches new grants from Pro to Plus

The program now targets 200,000 researchers. Applications for 50,000 new grants open November 23, with first-come selection and Plus access for recipients and up…

Logo oficial da Microsoft sobre foto de cibersegurança, capa da notícia: Microsoft lança MAI-Cyber-1-Flash, modelo de IA que bate rivais em cibersegurança pela metade do preço
137 · Releases & Models

Microsoft Launches MAI-Cyber-1-Flash, an AI Model That Beats Rivals on Cybersecurity at Half the Price

Microsoft launched MAI-Cyber-1-Flash on July 27, 2026, its first cybersecurity specialized AI model, alongside the Project Perception platform. Running inside MDASH, the model scored…

Logo oficial da Kimi sobre foto de inteligência artificial, capa da notícia: Kimi K3 vira o maior modelo de IA de peso aberto do mundo com pesos liberados de graça
138 · Releases & Models

Kimi K3 Becomes the World's Largest Open Weight AI Model as Moonshot Releases It for Free

Moonshot AI released the open weights for Kimi K3 on July 27, 2026, a 2.8 trillion parameter model that becomes the largest open weight…

Logo oficial da Mistral AI sobre foto de inteligência artificial, capa da notícia: Mistral confirma nova família de modelos abertos “gorda mas esparsa” com acesso antecipado em julho
139 · Releases & Models

Mistral Confirms a New "Fat but Sparse" Open Model Family, With Early Access Starting in July

Mistral AI CEO Arthur Mensch confirmed a new open weight model family using a mixture-of-experts architecture he describes as "fat but sparse," with early…

Logo oficial da OpenAI sobre foto de saúde, capa da notícia: OpenAI libera o ChatGPT Health para todos os adultos dos EUA um dia depois de processo pedir bloqueio
140 · Releases & Models

OpenAI Opens ChatGPT Health to All US Adults One Day After a Lawsuit Sought to Block It

OpenAI opened ChatGPT Health to every adult in the United States on July 23, 2026, one day after a former Florida pastor asked a…

Logo oficial da Anthropic sobre foto de inteligência artificial, capa da notícia: Anthropic lança Claude Opus 5 com mesmo preço da geração anterior e desempenho perto do Fable 5
141 · Releases & Models

Anthropic Launches Claude Opus 5 at Same Price as Previous Generation, Nearing Fable 5 Performance

Anthropic launched Claude Opus 5 on July 24, 2026, a more cost efficient model for everyday tasks that keeps the same pricing as Claude…

Logo oficial da Gemini sobre foto de inteligência artificial, capa da notícia: Google lança trio de Gemini Flash, mas o 3.5 Pro segue sumido e o Gemini 4 já começou a treinar
142 · Releases & Models

RUMOR: Google Reportedly Has Gemini 4 Pro's First Checkpoint Running, Eyeing an October Launch

An AI insider on X claims Google has already run Gemini 4 Pro's first internal checkpoint and is targeting an October public launch, with…

Logo oficial das marcas xAI e Kimi sobre foto de inteligência artificial, capa da notícia: Musk anuncia Grok 4.6 com 2 trilhões de parâmetros pra responder ao Kimi K3
143 · Releases & Models

Musk Announces 2 Trillion Parameter Grok 4.6 to Answer Kimi K3

On July 18, 2026, Elon Musk revealed that xAI is training Grok 4.6, a 2 trillion parameter model expected to finish its first training…

Logo oficial das marcas Qwen e Claude sobre foto de inteligência artificial, capa da notícia: Alibaba mostra o Qwen3.8-Max e diz que o modelo fica atrás só do Claude Fable 5
144 · Releases & Models

Alibaba Previews Qwen3.8-Max, Says the Model Trails Only Claude Fable 5

On July 19, 2026, during WAIC in Shanghai, Alibaba unveiled Qwen3.8-Max-Preview, a 2.4 trillion parameter multimodal model the company positions as the world's second…

Ilustração editorial de um smartphone com tela exibindo uma rede de agente de IA conectando ícones de aplicativos, flutuando no palco de uma conferência de tecnologia em Xangai, referência ao lançamento do NaviX Ultra da ZTE/Nubia
145 · Releases & Models

ZTE Launches NaviX Ultra, the World's First Smartphone With an AI Agent Built Into the OS

ZTE unveiled the NaviX Ultra at WAIC in Shanghai, a Nubia sub-brand phone running ByteDance's Doubao assistant that promises to complete tasks across multiple…

Logo oficial da Gemini sobre foto de inteligência artificial, capa da notícia: Google atrasa Gemini 3.5 Pro por meses após novo retrabalho no treinamento
146 · Releases & Models

Google Delays Gemini 3.5 Pro Launch by Months After Retraining Falls Short

Google has once again pushed back the broad rollout of Gemini 3.5 Pro, promised for June 2026 by Alphabet CEO Sundar Pichai at Google…

Capa editorial Max IA Assistant com o tema Thinking Machines · Inkling sobre foto de inteligência artificial, notícia: Mira Murati lança o Inkling, primeiro modelo aberto da Thinking Machines, e admite que não é o mais forte do mercado
147 · Releases & Models

Mira Murati Launches Inkling, Thinking Machines' First Open Model, and Admits It's Not the Strongest on the Market

Thinking Machines, Mira Murati's startup valued at $12 billion, launched Inkling on July 15, 2026, a 975 billion parameter open model the company itself…

Logo oficial da Kimi sobre foto de publicidade e preços, capa da notícia: Kimi K3, da Moonshot AI, chega como o modelo aberto mais avançado do mundo e custa fração do preço
148 · Releases & Models

Moonshot AI's Kimi K3 Arrives as the World's Most Advanced Open Model, at a Fraction of the Price

Moonshot AI launched Kimi K3 on July 16, 2026, a 2.8 trillion parameter model that ranked fourth on the Artificial Analysis Intelligence Index, priced…

Logo oficial da xAI sobre foto de inteligência artificial, capa da notícia: SpaceXAI anuncia Grok 4.5, modelo “classe Opus” mais barato que a concorrência
149 · Releases & Models

SpaceXAI announces Grok 4.5, an "Opus-class" model cheaper than rivals

On July 8, 2026, SpaceXAI (formerly xAI) announced Grok 4.5, a model the company describes as comparable in class to Claude Opus 4.7, with…

Rede neural abstrata em tons coral representando o lancamento do Claude Sonnet 5 pela Anthropic
150 · Releases & Models

Anthropic launches Claude Sonnet 5, its most agentic Sonnet model yet

Anthropic launched Claude Sonnet 5 on June 30, 2026, now the default model on the Free and Pro plans, with a 1 million token…

Logo da Anthropic sobre fundo gradiente azul-roxo, capa da notícia sobre o lançamento do Claude Fable 5 e Mythos 5
151 · Releases & Models

Anthropic launches Claude Fable 5 and Mythos 5

Anthropic launched two frontier models on June 9: Fable 5, the public version with a reinforced safety layer, and Mythos 5, which shares the…

Logo da xAI sobre fundo gradiente azul-roxo, capa da noticia do Grok V9-Medium
152 · Releases & Models

xAI launches Grok V9-Medium, a coding-focused model

Grok V9-Medium, a 1.5-trillion-parameter model trained on real Cursor session data, arrived on Grok and SuperGrok on June 16.

Logo da OpenAI sobre fundo gradiente azul-roxo, capa da noticia do GPT-5.6
153 · Releases & Models

OpenAI previews GPT-5.6 with restricted access

OpenAI announced a limited preview of the GPT-5.6 family (Sol, Terra and Luna) on June 26, currently available to only about 20 companies approved…

Logo da LangChain sobre fundo gradiente azul-roxo, capa da noticia do Deep Agents e LangGraph 1.2.4
154 · Releases & Models

LangChain launches Deep Agents and updates LangGraph to version 1.2.4

LangChain launched Deep Agents in March, a higher-level framework built on LangGraph, while LangGraph itself reached version 1.2.4 on June 2.