Google is preparing Interactive Reports for Gemini Notebook, a feature that embeds Studio content into customizable reports, as indicated by a new notice in the Studio panel. The feature, described as…
OpenAI is moving its next frontier model, reportedly codenamed 'mozaik-alpha-fdm' and possibly named GPT-6 'Astra,' closer to release, with expanded internal testing and outputs showing strong coding …
Tencent released Hy4 preview, a 770-billion-parameter open-source flagship model with 49 billion active parameters and a one-million-token context window, targeting long-horizon software engineering, …
Meta is preparing to launch Hatch, an autonomous consumer agent that performs online tasks such as filling out forms, making purchases, and booking restaurants, with unreleased materials revealing a b…
Z.ai launched GLM-5.3-Flash, the first natively multimodal model in its GLM-5 family, under an MIT license, featuring 320 billion total parameters and 18 billion active parameters. The company claims …
Anthropic is rolling out a unified Claude memory that follows users between chat and Claude Cowork, starting today for Free, Pro, and Max users on web, desktop, and mobile, with Team and Enterprise ac…
Google released Antigravity 2.0, adding Git version-control controls and an embedded terminal to its coding assistant's right-side panel, enabling developers to review repository changes and run comma…
NVIDIA announced that SpaceXAI will deploy its Vera CPUs to power agentic AI workloads behind Grok, with plans to extend the Vera Rubin NVL72 platform into orbit via the Starmind AI satellite. NVIDIA …
Google is preparing its Gemini desktop app for avatars, plugins, and the upcoming Gemini 4 model, with avatars and a hidden Customize area for apps, skills, and plugins appearing in the latest version…
Meta is beta-testing its upcoming Muse Video model with selected partners, and early outputs show state-of-the-art potential in fine detail, world understanding, and temporal consistency, producing 10…
Meta launched its Meta AI desktop app for macOS on August 19, featuring screen sharing and system-wide dictation, with the build already labeled version 1.0. The app allows users to attach a specific …
Fabraix has opened public access to Nyx, its autonomous red-teaming agent for AI security, which tests chat, voice, browser, and coding agents using a library of over 10,000 jailbreaks and attack stra…
DeepReinforce released Ornith-1.5, a family of open models in 397B, 35B, and 9B sizes, extending the self-scaffolding framework into a closed self-improvement loop where the system proposes tasks, gen…
Google is testing computer-use functionality in its Gemini desktop app, which would allow the AI to control other apps, access files, and perform tasks similar to Claude and ChatGPT. The feature, expe…
Anthropic is internally testing a new feature called Hub Mode in its Claude mobile app, which appears to be a revamped version of its previously removed Orbit feature and is designed to manage sub-age…
Anthropic is developing collaborative projects for Claude Code that add Slack-like features, including persistent repositories as context, multiple threads within sessions, and shared memory across se…
Anthropic is developing a built-in web browser for Cowork sessions in Claude Desktop, according to TestingCatalog, which would allow users to browse the web while Claude maintains context, bridging a …
Google has launched Sheets canvas, a Gemini-powered feature in Google Sheets that lets users create interactive mini-apps from spreadsheet data using natural language, with real-time two-way syncing. …
Google launched Gemini 3.7 Flash, its most intelligent workhorse model for coding and agents, three weeks after Gemini 3.6 Flash. The model scored 43.6% on FrontierCode 1.1 Main (up from 34.4%), 65.3%…
OpenAI has opened a limited preview of Ultrafast, a new API service tier that runs GPT-5.6 Sol at up to 14 times the speed of Standard processing, generating up to 750 output tokens per second. Powere…