Posts in AI （738）

NVIDIA Begins Vera Rubin NVL72 Production Across 350 Sites, Powering Mistral's European AI Infrastructure

OpenAI Launches ChatGPT for Small Business with Training, Guides and Partner Perks

NVIDIA Spectrum-6 Reaches AI Factories: 102.4 Tb/s Doubles Bandwidth for Vera Rubin

Google Ships Gemini 3.6 Flash with 17% Fewer Output Tokens and a Lower Price, Alongside 3.5 Flash-Lite

* Image is for illustration purposes.

Alibaba Previews Qwen3.8-Max, a 2.4 Trillion-Parameter Model, With Open Weights Still Undated

Google Reportedly Developing Frozen v2, a Chip With Gemini's Architecture Etched Into Silicon

* Image is for illustration purposes.

Grok for Excel Arrives: xAI's AI Handles Spreadsheet Work and Cites the Cells It Used

* Image is for illustration purposes.

Sakana AI Releases Defense-Focused Fugu-Cyber, Scoring 86.9% on CyberGym

AMD Ships Its First Rack-Scale AI System Helios as Microsoft Commits Azure at Scale

* Image is for illustration purposes.

OpenAI Publishes Safety Report on Long-Running Models After Sandbox Escape Forced a Pause

* Image is for illustration purposes.

Claude Fable 5 Becomes Standard on Max and Team Premium as Pro Moves to Pay-As-You-Go

Anthropic Opens Rare Disease Research Grants with Up to 50,000 USD in Claude Credits, Applications Close August 2

NVIDIA Pushes Agentic AI Into Creative Tools at SIGGRAPH, With Synthetic Video Detection and Cosmos 3 Edge

OpenAI's First Hardware Is Said to Be a Screenless Speaker Built to Feel Alive

* Image is for illustration purposes.

Moonshot AI Releases Kimi K3: A 2.8-Trillion-Parameter Open Model Closing In on GPT and Claude

* Image is for illustration purposes.

Apple Intelligence Cleared in China, Backed by Alibaba's Qwen and Baidu

India's Cars24 Rolls Out OpenAI Voice and Chat Agents, Automating Over 1M Monthly Conversation Minutes

NVIDIA Frames Intelligence per Dollar as a New AI Metric, With Vera Rubin Training Top Models on One-Fourth the GPUs

* Image is for illustration purposes.

OpenAI Unveils Automated Red-Teamer GPT-Red, Hardening GPT-5.6 Against Prompt Injection

NVIDIA Says Performance per Watt Is the Key Metric for AI Factories, With Blackwell Delivering Up to 25x

Anthropic Launches Claude for Teachers, Giving US K-12 Educators Free Premium Access

Meta to Start Mass Production of Its Custom 'Iris' AI Chip in September, Cutting Nvidia Reliance

Google Details Gemini's Rise in Southeast Asia: 70% Native-Language Prompts, Users Double in a Year

* Image is for illustration purposes.

Meta Launches Muse Spark 1.1 Coding Model and Its First Paid API to Rival OpenAI and Anthropic