|
Saturday, May 30, 2026
|
|
|
|
| |
|
Today’s top story
FUNDING
Anthropic raises $65B Series H at $965B valuation, tops OpenAI
TechCrunch
Anthropic closed a $65 billion Series H led by Altimeter, Dragoneer, Greenoaks and Sequoia at a $965 billion post-money valuation, leapfrogging OpenAI's $730 billion mark to become the most valuable private AI company. The round includes $15 billion in previously committed hyperscaler money, with Amazon adding $5 billion and infrastructure partners Micron, Samsung and SK hynix joining. Anthropic disclosed run-rate revenue had crossed $47 billion earlier in May.
|
|
Why it matters
The raise resets the pecking order at the frontier, giving Anthropic the balance sheet to keep buying compute at OpenAI scale and accelerating an apparent IPO timeline. It also locks memory and storage suppliers into Claude's roadmap, deepening the chip-supply moats that increasingly determine who can train next-generation models.
|
|
|
More headlines
|
|
FUNDING
Groq lines up $650M for inference neocloud after Nvidia's $20B raid
Groq is raising $650 million from existing investors led by Disruptive and Infinitium to fund a pivot toward an inference-focused neocloud business built on its homegrown LPU chips. The round follows December's $20 billion not-an-acquisition with Nvidia, which absorbed senior Groq engineers and licensed its hardware while leaving the company to chart a second act under interim CEO Adam Winter.
TechCrunch
|
|
RELEASE
Google ships Gemini 3.5 Flash across Antigravity and the Gemini API
Google made Gemini 3.5 Flash generally available through Antigravity, AI Studio and Android Studio, claiming it beats Gemini 3.1 Pro on coding and agentic benchmarks at Flash-tier latency and price. The release headlines a broader I/O 2026 push to keep frontier capability inside cheaper, faster models that can be embedded into consumer products at scale.
Google Blog
|
|
INDUSTRY
Anthropic in talks to run Claude inference on Microsoft's Maia 200 chips
Anthropic and Microsoft are in early discussions to port portions of Claude's inference workload onto Microsoft's second-generation Maia 200 accelerator on Azure, with technical evaluations running first before any commercial terms. A deal would be Maia's first production test serving a frontier model it didn't build, and would give Anthropic a route around Nvidia capacity constraints.
CNBC
|
|
RELEASE
OpenAI opens GPT-Rosalind to vetted biodefense developers and US partners
OpenAI launched Rosalind Biodefense, extending trusted access to its GPT-Rosalind model for screened developers and U.S. government partners working on pandemic preparedness, public health and biosecurity. The program formalizes a gated-deployment pattern OpenAI has been piloting for dual-use scientific models that the company does not want broadly exposed.
OpenAI
|
|
INDUSTRY
Dell posts $60B in AI server sales as Micron crosses $1T market cap
Dell's first-quarter results showed roughly $60 billion in AI server sales, sending the stock up as much as 40% on signs that enterprise GPU deployments are still accelerating. Memory supplier Micron crossed a $1 trillion market capitalization for the first time this week on AI-driven demand, up 29% in five sessions.
Yahoo Finance
|
|
This week in benchmarks
|
|
vs. Thursday, May 21, 2026
|
|
Intelligence
GPT-5.5 (xhigh) fell from #1 to #2
|
|
Intelligence
GPT-5.5 (high) fell from #2 to #3
|
|
Intelligence
Claude Opus 4.7 (Adaptive Reasoning, Max Effort) fell from #3 to #4
|
|
Intelligence
Gemini 3.1 Pro Preview fell from #4 to #5
|
|
Intelligence
GPT-5.4 (xhigh) fell from #5 to #6
|
|
Intelligence
GPT-5.5 (medium) fell from #6 to #7
|
|
Read past briefs and the live benchmarks at
ai-newsletter-eta-lemon.vercel.app.
Forward this to a friend who’d appreciate the brief.
|
