Anthropic priced its IPO at $83 per share — the largest in U.S. history — and begins trading on the Nasdaq today. Google launched Gemini 4 Neon, and Korean startup FuriosaAI began mass production of its RNGD inference chip.
📈 Anthropic IPO Priced at $83… Largest in U.S. History, Trading Starts Today on Nasdaq
Anthropic priced its initial public offering at $83 per share, above its $70 to $78 range. At about $81.3 billion in total gross proceeds, it is the largest IPO in U.S. history, according to Renaissance Capital, with a fully diluted valuation of roughly $1.5 trillion. Shares begin trading on the Nasdaq today under the ticker ANT, giving investors their first chance to buy into one of OpenAI's biggest rivals.
✨ Google Launches Gemini 4 Neon… Mid-Tier Model for Coding and Complex Analysis, 1M-Token Context
Google launched Gemini 4 Neon, the mid-tier model in the Gemini 4 family, positioned between Carbon for the hardest long-horizon tasks and Hydrogen for fast everyday work. Aimed at high-frequency complex coding, analytical reasoning and enterprise workflows, it is priced at $3 per million input tokens and $15 per million output tokens, with a 1-million-token context window, available through the API and the Gemini app. Google also opened its interactive AI world model Genie 4 to more AI Ultra subscribers.
🤖 Mixed Early Reviews for GPT-6.2… Testers Criticize Autonomy and Usability
OpenAI's GPT-6.2 Astra and Sol, released earlier this week, drew criticism from some early testers over autonomy and usability. On the Artificial Analysis Intelligence Index, however, Astra scored 62, second only to Claude Fable 5.2 at 65, while Sol scored 57. OpenAI also said in an internal study that its coding agents now complete 46% of researchers' two-to-four-hour tasks, up from 23% in March.
🚫 NVIDIA Halts AI Chip Shipments to China… Despite U.S. Approval for Some Sales
NVIDIA has stopped shipping certain AI chips to China, even after Washington cleared some of the sales. Which chips are affected and how long the halt will last remain unclear. The move adds fresh uncertainty around AI chip export controls amid U.S.–China tech tensions, complicating the calculus for chipmakers eyeing the Chinese market.
🔧 OpenAI Uses AI Models to Improve Chip Design… Applied to Jalapeño Silicon and AMD GPUs
OpenAI said its AI models are now being used to improve chip designs, delivering measurable gains in its own Jalapeño inference chip and AMD's flagship GPU designs. The virtuous cycle of AI designing AI hardware is gathering pace, deepening collaboration between model developers and chipmakers.
🇰🇷 FuriosaAI Begins Mass Production of RNGD Inference Chip… 2x Power Efficiency vs NVIDIA's Top Chip
FuriosaAI, the Korean startup led by former Samsung executive June Paik, has begun mass production of its next-generation AI inference chip RNGD (Renegade) this month. The chip is rated at more than twice the power efficiency of NVIDIA's top chip and has passed technical validation from LG AI Research and OpenAI. The company was recently valued at about $700 million in fundraising and is reported to have turned down an acquisition offer from Meta last year.
🏥 KT and Seoul National University Hospital Trial Medical AI Robot for Child Patients
KT and Seoul National University Hospital have begun a pilot of a medical physical-AI robot for child patients. The robot talks with young patients, explains examination procedures and walks with them to destinations inside the hospital. Meanwhile, Samsung Electro-Mechanics is expanding package-substrate supply for NVIDIA's next-generation GPUs and adding MLCC cooperation, while SK Telecom is teaming up with Hyosung Heavy Industries to build gigawatt-scale AI data centers, as Korean Big Tech's AI infrastructure push continues.