Zuckerberg Sets ‘Aggressive’ Price Strategy for Meta’s Paid AI Tools
On July 9, local time, Meta dropped its most advanced AI model yet, Muse Spark 1.1, and for the first time, it’s rolling out a paid version for developers. This marks Meta’s first move to charge businesses for model access. In an interview before the launch, Zuckerberg said this model would be one of the cheapest options on the market. Meta is also introducing a new Meta Model API system to bill developers. He noted that the API pricing is only about 25% of what OpenAI and Anthropic charge for their top-tier models. Developers can still use Meta’s models for free, but once they hit a certain token usage threshold, they’ll need to pay up.
Samsung Develops GAIA AI PC Accelerator Chip, Sends Samples to Lenovo and HP for Testing
On July 10, reports surfaced that Samsung Electronics is cooking up an AI PC accelerator chip called GAIA. They’ve already shipped prototypes to Lenovo and HP for performance testing, with plans to start mass production next year. Built on a 4nm process, GAIA packs a neural processing unit (NPU) designed for generative AI tasks and supports processing-in-memory (PIM) technology.
Shang Wei New Materials to Unveil First Deformable Personal Robot Qiyuan T1 at WAIC 2026
On July 10, NUPIAO learned that Shang Wei New Materials will launch its first deformable personal robot, Qiyuan T1, at WAIC. This gadget can automatically switch between a wheeled humanoid and a four-legged form, all from one base design. It also packs camera tracking features and makes big leaps in smart interaction and companionship. Word has it that Qiyuan T1 breaks away from the typical humanoid robot mold, using its shape-shifting design to fit a bunch of home scenarios, redefining what a personal robot can be for the masses.
AI-Native Kimi Credit Card Locks in American Express and Agricultural Bank of China
On July 10, according to National Business Daily, the world’s first AI-native credit card, the “Kimi Credit Card,” has officially partnered with American Express and Agricultural Bank of China. It’s set to launch publicly today.
Here’s the deal: cardholders can use their everyday spending points to snag Kimi membership perks at a 20% discount. The points-to-cash ratio in the rewards mall is 1000:1, with a max of 50% of a purchase covered by points. Down the road, the three partners plan to explore more possibilities like AI agent payments, linking token allowances to credit limits, and swapping points for computing power.
OpenAI Unleashes ChatGPT Work: A GPT-5.6-Powered Work Agent
On July 9, local time, OpenAI dropped a brand-new enterprise feature: ChatGPT Work. This smart agent, powered by the cutting-edge GPT-5.6 model, isn’t your average chatbot. Designed for complex, multi-step projects, ChatGPT Work lets you throw in a final goal, and it’ll break it down, plan it out, pull context from connected tools, and auto-generate documents, spreadsheets, or presentations. Plus, the newly launched ChatGPT Sites feature can turn your ideas or data into lightweight collaborative websites with a single click. Right now, it’s live on macOS and Windows desktops, rolling out gradually to paying users like Plus and Enterprise plans. And don’t worry—users and managers still have full control and approval power over the whole process.
Google Mandates AI Ad Labels Across Search, YouTube, and More
On July 9, local time, Google announced a big expansion in how it discloses generative AI ads. New rules now require clear labels on commercial ads created or edited by AI across Google Search, YouTube, and the Discover feed. The goal? To keep consumers from being fooled by fake, AI-generated scenes.
Musk Admits Underestimating Anthropic, Says He Won’t Cut Off Support Despite Competition
On July 9, Elon Musk took to social media to admit he was “clearly wrong” about AI company Anthropic, calling it “the clear leader in AI right now.” He added, “No company has released a model as good as Mythos/Fable yet.” Even though Anthropic is a competitor, Musk said he wouldn’t take steps that would seriously hurt the company. This came after a social media user pointed out that Anthropic relies on computing resources leased from SpaceX and speculated about the fallout if SpaceX stopped that service. Musk responded by saying he wouldn’t pull such a move.
MiniMax Closes $2 Billion+ Funding Round
On July 10, MiniMax wrapped up a new funding round worth 16 billion Hong Kong dollars (over $2 billion).This deal attracted a mix of top-tier international sovereign funds, long-term investors, leading Chinese institutions, and elite multi-strategy funds from across Asia-Pacific, Europe, and the U.S. Among them were over 20 long-term and sovereign funds, with several pre-IPO and cornerstone investors upping their stakes in this round.Reports say the round drew interest from over 100 institutions, with 7x oversubscription. Initially set at about $1.8 billion, the final tally grew to over $2 billion thanks to strong institutional demand.
