Google made Gemini Tasks free in the US, Japan, Korea, Canada, Australia and India on November 17, putting a calling, paying agent in front of hundreds of millions — while its payment rail, not its model, does the locking in.
Alibaba says Qwen4-Max cuts average cost per task by about 60% against its predecessor. Independent testing confirms the economy and denies the supremacy: 43 on Artificial Analysis Intelligence Index, $3.27 per task, 32 minutes per task, 17 points behind Gemini 4 Carbon.
OpenAI launched ChatGPT Wallet on November 13, 2026 with American Express, Mastercard and Visa: dots agents can pay for purchases, with a biometric check on every transaction and network tokens replacing real card numbers.
Google released Gemini 4 Hydrogen, its cheapest Gemini 4 model, at $1.50 per million input tokens and $7.50 per million output tokens, scoring 57 percent on Terminal-Bench 4.1 and cutting fabricated claims five-fold while serving free users through auto model selection.
OpenAI's Q3 revenue reportedly crossed $10 billion for the first time, up nearly 50% from the prior quarter, even as its adjusted operating loss widened to $10.9 billion.
OpenAI's beta Businesses in ChatGPT lets companies claim verified profiles, sync product feeds, and deploy Business Agents for its 1.2 billion weekly users - a direct challenge to Google's discovery-and-ads economy.
Google opened Gemini 4 Carbon to all paying developers on 10 November and previewed /swarm, a multi-agent mode in Antigravity, at $2/$10 per million tokens. Independent scoring puts it fourth at 60, with $7.76 and 24.9 minutes per index task.
Google DeepMind's AlphaProof Code split the work between people and agents to formally verify Signal's shipped implementation: over six million lines of Lean in about four weeks, with humans reviewing roughly one percent of the codebase.
Anthropic's 6 November post scores agent memory on CL-Bench 1.0: Opus 5.5 rises from 0.655 to 0.838, yet a rule to record only failures cuts the gain by 0.234. Storage size is not the point.
Baidu released ERNIE 5.5 Flash Base: 280B parameters, 14B active per token, 33.5 trillion training tokens, FP8 and BF16 weights. No chat tuning, no permissive licence, and a benchmark table Baidu ran on itself.
OpenAI's cyber model proved 948 local privilege-escalation bugs in the Linux kernel for $1.2 million. Ten days later only 61 were patched upstream. Finding is now cheaper than fixing.
SemiAnalysis tested 13 frontier models on 206 real professional tasks. The winner finished 12% of them, at about $51 a task. The same week, Anthropic's filing showed what that arithmetic does to a business.
Page 1 of 2 ·
18 posts
About this feed
How often is the News page updated?
+
The feed is updated whenever there is something worth reporting: AI industry announcements, and changes to the AI Bookmarks directory itself. Every article carries its publication date, and the whole feed is mirrored in the RSS file.
Can I get these updates by email instead?
+
Yes. The Email updates page subscribes you to a digest of new articles. Every email links straight back to the article on this site and carries a one-click unsubscribe link.
How do I suggest a story or a tool we have missed?
+
Use the Submit a tool page for tools, or the Contact form for anything else. Suggestions are reviewed before they appear.
Get in touch
Tell us what you need and we will reply by email, usually within five working days.
✓
Thank you!
Your message has been sent. We will get back to you by email soon.