Archive

Today's Key Insights

  • Anthropic Revenue Tops $65B; Fall 2026 IPO Could Beat OpenAI — Anthropic’s possible fall 2026 listing matters to public-market investors because Bloomberg places a $1 trillion valuation on the deal and says it could beat OpenAI to market.
  • Stripe Reportedly Acquiring OpenRouter for More Than $7B — The reported acquisition would put OpenRouter's eight million users and access to more than 400 AI models under Stripe, giving Stripe control of an AI gateway valued at more than five times its latest reported valuation.
  • Groq Raises $350M as AI-Chip Maker Pivots to Neocloud — Groq raised $350 million to fuel its pivot from AI chips to a neocloud business while expanding its Nvidia-powered data-center footprint.
  • Zhipu Says GLM-5.3 Improves Fastest in Its Weakest Area — Security teams assessing Zhipu’s GLM-5.3 results should read the claim as progress from Zhipu’s weakest area—not as a statement that the model leads in cybersecurity.
  • Z.ai’s Latest Model Could Aid Defenders—or Hackers — Wired AI frames Z.ai’s latest model as a security tool for companies—and a security risk if hackers obtain it.

Top Story

Anthropic Revenue Tops $65B; Fall 2026 IPO Could Beat OpenAI #

The Decoder AI • TechCrunch AI

Anthropic’s annualized revenue has topped $65 billion, a sevenfold increase in one year, according to Bloomberg. TechCrunch reported that the Claude maker added $18 billion to its annualized revenue in two months.

Bloomberg reported that Anthropic could go public as early as fall 2026 at a $1 trillion valuation, potentially beating OpenAI to market.

Why it matters: Anthropic’s possible fall 2026 listing matters to public-market investors because Bloomberg places a $1 trillion valuation on the deal and says it could beat OpenAI to market.

Key Takeaways

  • Bloomberg reported that Anthropic’s annualized revenue grew sevenfold in one year.
  • TechCrunch reported an $18 billion addition to Anthropic’s annualized revenue over two months.
  • The reports describe the $65 billion figure as annualized revenue.

Industry Updates

Stripe Reportedly Acquiring OpenRouter for More Than $7B #

The Decoder AI • TechCrunch AI

Stripe is reportedly acquiring AI gateway startup OpenRouter for more than $7 billion, according to Bloomberg. The reported price is up sharply from OpenRouter's latest $1.3 billion valuation.

OpenRouter offers access to more than 400 AI models and has eight million users. Its CEO recently described the startup as “Stripe for AI.”

Why it matters: The reported acquisition would put OpenRouter's eight million users and access to more than 400 AI models under Stripe, giving Stripe control of an AI gateway valued at more than five times its latest reported valuation.

Groq Raises $350M as AI-Chip Maker Pivots to Neocloud #

TechCrunch AI

Groq raised $350 million at a $3.5 billion valuation as the former AI chipmaker pivots to a neocloud business. TechCrunch AI reports that Groq is also expanding its Nvidia-powered data-center footprint.

The financing and infrastructure expansion mark Groq's shift from its earlier identity as an AI chipmaker toward a neocloud operation.

Why it matters: Groq raised $350 million to fuel its pivot from AI chips to a neocloud business while expanding its Nvidia-powered data-center footprint.

Zhipu Says GLM-5.3 Improves Fastest in Its Weakest Area #

AI News

Zhipu’s GLM-5.3 release note says its cybersecurity capability “is growing fastest exactly where we are furthest behind.” The Beijing-based company, which also trades as Z.ai, included the sentence in a release note that AI News says did not appear in most coverage.

That wording puts GLM-5.3’s cybersecurity result in a specific frame: Zhipu describes the fastest improvement in the area where it says the lab remains furthest behind. Zhipu is one of a handful of Chinese labs releasing competing models.

Why it matters: Security teams assessing Zhipu’s GLM-5.3 results should read the claim as progress from Zhipu’s weakest area—not as a statement that the model leads in cybersecurity.

Z.ai’s Latest Model Could Aid Defenders—or Hackers #

Wired AI

Z.ai has released its latest AI model, which Wired AI describes as a powerful Chinese model that experts warned about and awaited.

The report gives the release a dual-use framing: companies could use the model to secure their systems, but it could also find its way into hackers’ hands. Wired AI provides no further details about the model’s capabilities or availability.

Why it matters: Wired AI frames Z.ai’s latest model as a security tool for companies—and a security risk if hackers obtain it.