Three of AI's biggest companies asked the public to trust their own numbers today. Nvidia reported earnings for the first time since Jensen Huang quit. Google and Anthropic each advertised a cheaper, faster model.
LVMH handed its newsrooms to OpenAI, betting readers will trust a machine summary the way they trust a byline. Each move asks the same question in a different market.
Nvidia Reports Alone for the First Time
Colette Kress runs today's third-quarter call by herself1. It is her first call since Huang quit on Nov. 12. Wall Street wants $108 billion in revenue, a jump near 90 percent from a year ago. A founder's exit rarely shows up in a chip order placed months earlier.
The stock has fallen 35 percent from its October peak, erasing more than a trillion dollars in value2. Tonight's number settles whether that drop tracked Huang's exit or the chips themselves.

Google Says Its Cheap Model Reaches Everyone
Gemini 4 Hydrogen shipped to everyone today3. "Today, we're introducing Gemini 4 Hydrogen, bringing cutting-edge intelligence to everyone — with speed and efficiency for everyday coding and agentic workflows," wrote Koray Kavukcuoglu, Google DeepMind's chief AI architect.
The model is fast3. It finishes everyday coding tasks two and a half times faster than Google's last Flash model. Hydrogen also triples that older model's score on Terminal-Bench, a timed test of real command-line work.
Artificial Analysis tested it too4. The outside firm scored Hydrogen at 54 percent on that terminal test, behind Claude Sonnet 5.5's 64 and GPT-6.1 Sol's 56. Speed improved first, and terminal skill lagged behind it.
Artificial Analysis@ArtificialAnlysGoogle's new Gemini 4 Hydrogen has Intelligence just below Claude Sonnet 5.5 and GPT-6.1 Sol, at a third of Sonnet’s cost but ~2.5x that of GPT-6.1 Sol@GoogleDeepMind has released their third new model in six weeks. Gemini 4 Hydrogen introduces an xhigh reasoning level and… pic.twitter.com/mtn4GzkVVq
Hydrogen Scores 10 Points Behind Sonnet on the Terminal Test
Terminal-Bench 4.0 score, percent of command-line tasks solved, each model at its top reasoning setting; higher is better
Each column is one model's score on Artificial Analysis's terminal coding test; the magenta column is Hydrogen, the model Google shipped today.
Source: Artificial Analysis [4]. Chart by Ryan Elliott Dennis.
The numbers behind this chart
| Item | Value |
|---|---|
| Claude Sonnet 5.5 | 64% |
| GPT-6.1 Sol | 56% |
| Gemini 4 Hydrogen | 54% |
Anthropic's Price Cut Meets a Skeptic
Anthropic cut Fable 5.2's price by a fifth, 20 percent, on Nov. 95. Dianne Penn, the company's head of product management for research, spoke with CNBC about the move. She said it answers Dario Amodei's September call for a slower pace. "We called for pacing, not pausing," she said.
Endor Labs ran matching coding tasks on both Fable versions and found the real bill moved further than the price did6. Fable 5.2's median cost per task fell 77 percent, to 54 cents. Output tokens fell 68 percent.
Fable 5.2's Real Savings Nearly Quadruple Its List Price Cut
Percent drop from Fable 5.1 to Fable 5.2: Anthropic's list price per token against Endor Labs' measured cost per finished coding task
The first bar is the price cut Anthropic advertised; the second is the drop Endor Labs measured by running matching coding tasks on both models.
Sources: CNBC [5]; Endor Labs [6]. Chart by Ryan Elliott Dennis.
The numbers behind this chart
| Item | Value |
|---|---|
| Anthropic's list-price cut | 20% |
| Endor Labs' measured cost drop | 77% |
Sanchit Vir Gogia, chief analyst at Greyhound Research, read the cut a different way7. "Anthropic has cut the price of a token, not the price of a thought," he said. He urged buyers to measure their own cost per finished task, apart from the list price.
LVMH Hands Its Newsrooms to OpenAI
LVMH's media division signed a deal with OpenAI on Nov. 168. The deal covers 17 titles. Les Echos and Le Parisien are among them, and it puts ChatGPT Enterprise in front of their staff.
Michèle Benbunan, the division's chief executive, called the deal a step forward for her newsrooms. "This agreement will position our publications at the forefront of the media industry," she said8. Readers who ask ChatGPT about French news may now see those outlets' own summaries, with links back to the source. The deal treats a chatbot summary as publicity a newsroom can use.

MIT Gets One Model to Edit Another's Brain
Four MIT researchers posted a paper on Nov. 2 that coders still argue over9. Rikiya Takehi and his co-authors had OpenAI's GPT-6 Astra open a far smaller model, Qwen3.5-0.8B. The bigger model then edited the smaller one's own weights directly, skipping any training run. A model acting as its own engineer is still rare outside a research lab.
The edit touched just 0.016 percent of that model's weights9. It lifted the small model's math score from 7.9 percent to 48.1 percent. A grade-school benchmark climbed from 13.8 percent to 67.1 percent too.
A German Robot Maker Banks SoftBank's Money
Agile Robots closed a $500 million round on Nov. 12, led by SoftBank, with Nvidia joining in10. The Munich company's new valuation, $8.5 billion, makes it Europe's priciest private robotics firm today.
"Physical AI will fundamentally change how the world builds, moves and makes things," said founder and chief executive Zhaopeng Chen10. The money funds a push into the United States and more factory floor space back home. SoftBank's bet follows the same pattern it already placed on chips and robots elsewhere.
OpenAI Doubles Its Free Researcher Program
A free-access program for scientists doubled today, from 100,000 seats to 200,000 worldwide11. Since the program's July start, 20,000 researchers across 33 countries have already signed up.
New entrants get ChatGPT Plus, a step down from the Pro tier the first group received11. Each grant still covers one researcher. Two helpers can ride along too.
A Record Order Lands on Anthropic's Book
Brad Gerstner told CNBC on Nov. 16 that Altimeter Capital wants $5 billion of Anthropic's coming IPO12. Gerstner's order runs five times the largest one he has ever placed for any company going public.
"I think this is a cornerstone compounder for a long time," Gerstner said12. Seven chipmakers and cloud providers, Nvidia and Microsoft among them, are putting in $26 billion more at the same offer price. His order shows how badly Wall Street wants a piece of Anthropic before trading even starts.
Two Lines, One Each
Lovable and three other AI startups each reached $100 million in yearly revenue within eight months of launch13. An eight-month climb was rare for any software company before this year. Separately, California's frontier AI law met its first real test this fall14. OpenAI's GPT-6 Astra became the first model rated "Critical" under the state's new reporting rule.
What to watch
Watch Nvidia's own number tonight against the $108 billion guide, and whether Kress names a permanent successor on the call. See whether more buyers run their own task-level math on Fable 5.2 and Hydrogen, the way Endor Labs just did. Track whether LVMH's journalists get a say in how the OpenAI deal reaches their newsroom.
Sources
- Steven Porrello, "Nvidia Stock Earnings Preview: What to Watch on Nov. 17," The Motley Fool, Nov. 15, 2026, https://www.fool.com/investing/2026/11/15/nvidia-reports-earnings-on-nov-17-2026-heres-what-investors-should-be-watching/
- CNBC, "Nvidia stock drops for 6th week, shedding $1.6 trillion in market cap," CNBC, Nov. 13, 2026, https://www.cnbc.com/amp/2026/11/13/nvidia-stock-jensen-affair-ai.html
- Koray Kavukcuoglu, "Introducing Gemini 4 Hydrogen for Agentic Workflows," Google, Nov. 16, 2026, https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-hydrogen/
- Artificial Analysis, "Gemini 4 Hydrogen Models - Intelligence, Performance & Price Comparison," Artificial Analysis, Nov. 12, 2026, https://artificialanalysis.ai/models/gemini-4-hydrogen
- CNBC, "Anthropic releases Claude Fable 5.2 and Claude Mythos 5.2 ahead of IPO," CNBC, Nov. 9, 2026, https://www.cnbc.com/2026/11/09/anthropic-claude-fable-5-2-mythos-5-2.html
- Luca Compagna, "Fable 5.2: cheaper and faster, not more secure, and much more memorization," Endor Labs, Nov. 16, 2026, https://www.endorlabs.com/learn/fable-5-2-cheaper-and-faster-not-more-secure-and-much-more-memorization
- Prasanth Aby Thomas, "Anthropic unveils Claude Fable 5.2, Mythos 5.2 with coding gains and lower prices," InfoWorld, Nov. 10, 2026, https://www.infoworld.com/article/4246681/anthropic-unveils-claude-fable-5-2-mythos-5-2-with-coding-gains-and-lower-prices.html
- Content + Technology, "France's LVMH in Strategic Media Partnership with OpenAI," Content + Technology, Nov. 16, 2026, https://content-technology.com/ai-in-media/frances-lvmh-in-strategic-media-partnership-with-openai/
- Rikiya Takehi, Jayden Teoh, Alex Zhang, Omar Khattab, "Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights," MIT CSAIL, Nov. 2, 2026, https://rikiyat.github.io/blog/weight-space-programming/
- SiliconANGLE, "Agile Robots raises $500M funding led by SoftBank at $8.5B valuation," SiliconANGLE, Nov. 12, 2026, https://siliconangle.com/2026/11/12/agile-robots-raises-500m-funding-led-by-softbank-at-8-5b-valuation/
- IT Brief, "OpenAI expands free ChatGPT access for researchers," IT Brief, Nov. 17, 2026, https://itbrief.com.au/story/openai-expands-free-chatgpt-access-for-researchers
- CNBC, "Altimeter's Brad Gerstner aims to invest $5 billion in Anthropic IPO," CNBC, Nov. 16, 2026, https://www.cnbc.com/2026/11/16/altimeters-brad-gerstner-investing-5-billion-in-anthropic-ipo.html
- New Market Pitch, "Which AI startups have reached $100M ARR the fastest?," New Market Pitch, Nov. 10, 2026, https://newmarketpitch.com/blogs/news/ai-infrastructure-fastest-100m-arr
- Cloud Security Alliance, "CSA Research Note: SB 53 Frontier Disclosure Critical Threshold," Cloud Security Alliance, Nov. 10, 2026, https://labs.cloudsecurityalliance.org/research/csa-research-note-sb53-frontier-disclosure-critical-threshol/




