TechArtificial intelligence
Artificial intelligence

Alibaba prices Qwen4-Max at $1.50 per million tokens
Alibaba set its flagship model's rate at $1.50 per million input tokens and $4.50 per million output tokens in Singapore. The model page carries no benchmark results.
- Claude for Manufacturing drafts quotes and plant reports
Anthropic's new plugin has 17 skills and leaves drafts for a person to review. It doesn't run machinery or write to a company's systems.
- Google's Gemini 4 Hydrogen is rolling into Search and the app
The model replaces Gemini 3.8 Flash and costs $1.50 per million input tokens. The speed and performance numbers all come from Google.
- Four nonprofits sue to halt the most advanced AI
They want a San Francisco judge to order a stop to work on models more capable than today's at Anthropic, Google, Meta and OpenAI. Products already on the market are not targeted.
- Google launches a Gemini API that answers in probabilities
Gemini Decisions API scores up to 1,000 questions per request and charges $0.10 per million input tokens; the speed figures are Google's.
- Google brings Gemini Live to CarPlay for iPhone users
The voice assistant is rolling out worldwide, but Android Auto's proactive suggestions have no date and are headed for Pixel phones.
- Alibaba opens Qwen4-Max to subscribers only, through Nov. 25
Alibaba's new flagship model can't be paid for by usage and comes with no per-token price or test results; its capabilities are the company's own claims.
- OpenAI says Control Room halts research AI jobs in 35 seconds
Control Room covers the company's internal research, not ChatGPT or the API, and denies all inference if it goes down.
- Nvidia CEO Jensen Huang steps down, effective immediately
CFO Colette Kress takes over as interim CEO. The board gave no reason and will look for a permanent successor inside and outside the company.
- Claude can now build spreadsheets and whiteboards, in beta
Anthropic opened Sheets and Whiteboards to paid plans and took Claude Data out of beta. On Enterprise, the first two arrive switched off.
- ChatGPT will tell businesses how often it mentions them
OpenAI opened Business Hub in beta, a dashboard of 28-day mentions and clicks. Businesses do not see users' conversations.
- Acrobat is coming to Drive and Gmail, with no date or price
Google and Adobe announced an Acrobat shortcut in Drive and Gmail previews and Google models inside Photoshop. Neither company gave a date or a cost.
- Instagram lets users build AI Stories effects, US only for now
Effect Builder, a web tool inside Edits, turns a written description into an effect other people can try. Mexico is not covered yet.
- Google launches an AI agent for U.S. hospitals and insurers
Gemini for healthcare reads Epic and Oracle Health records without writing to them. It is a U.S.-only preview, and Google has published no accuracy or savings figures.
- Anthropic launches Claude Fable 5.2 and Mythos 5.2
Fable 5.2 is open to all users; Mythos 5.2, with fewer cybersecurity limits, is for verified organizations only.
- Google unveils two new AI chips, TPU 8tx and TPU 8i Max
One runs models on private data with protection built into the hardware; the other doubles memory for long contexts. No date, regions or price yet.
- None of ChatGPT's Pro plans has a five-hour limit, OpenAI says
An OpenAI staffer put numbers on each Pro tier: 10x, 20x and 50x Plus usage. The company still doesn't say what 1x is.
- Google's Gemini Omni 1.2 Flash video costs $0.07 a second
Google raised its clip limit to 15 seconds and cut the API price by 30%. The cheapest way to try it is Google Vids.

















