☕ Anthropic is raising $50B at $900B. Six months after hitting $380B.

Good morning, crowd!

Let me ask you something.

When was the last time you saw a company go from $380 billion to $900 billion in six months?

Not in a year. Not in a bull market over multiple quarters. In six months.

That's Anthropic. The AI company behind Claude just got offers for a $50 billion raise that would value it at $900 billion. For context, they were worth $380 billion in February. They hit $183 billion in September 2025. They started 2025 at $61 billion.

Do the math. They've grown 14x in 16 months.

Meanwhile, a four-month-old AI startup with 20 employees and no product just raised $500 million at a $4 billion valuation from Google and Nvidia. A corporate spend management company (yes, a fintech that handles expense reports) went from $16 billion to over $40 billion in less than a year. And Kraken just paid $600 million for a Hong Kong stablecoin payments firm you've probably never heard of.

This isn't normal. This isn't even 2021 levels of absurdity. This is something else entirely.

Welcome to Tuesday.

Let's break down what happened this week in private markets and what it means when valuations stop following logic and start following... something else. ☕

Jose, Alberto & the Founderscrowd team

In today's rundown:

Founderscrowd: Anthropic, the company behind Claude, is weighing offers for a $50 billion funding round at a $900 billion valuation. If it closes, Anthropic would surpass OpenAI ($852 billion) as the most valuable AI company in the world.

This isn't just big. It's unprecedented.

The numbers:

The revenue story:

Anthropic hit $30 billion in annualized revenue in early May 2026. Let me say that again because it's absurd: $30 billion ARR.

For context:

No American technology company has ever grown revenue this fast. Not Google. Not Facebook. Not Amazon. Nobody.

Enterprise customers now drive 80% of Anthropic's revenue. Over 1,000 businesses are spending more than $1 million annually on Claude. That's not consumer hype. That's real enterprise budgets getting reallocated to AI.

Why investors are scrambling:

Two catalysts converged:

1. Claude Code crushed it.

Anthropic's AI coding assistant became the go to tool for developers. The product works. Companies are paying for it. Revenue exploded.

2. Mythos changed the game.

On April 7, Anthropic unveiled Mythos, an advanced cybersecurity AI model. This sparked meetings between Trump administration officials, tech CEOs, and national security advisors.

When AI becomes a national security conversation, valuations stop being about profit multiples and start being about strategic importance.

The compute angle:

Anthropic secured 10 gigawatts of AI compute capacity in recent deals:

For context, 10 gigawatts could power 7.5 million homes. Anthropic is using it to train AI models.

The decision timeline:

Anthropic's board will decide in May 2026 whether to proceed with the raise. A potential IPO could come as early as October 2026.

Why this matters:

When a $380 billion company more than doubles its valuation in 90 days, one of two things is happening:

Both can be true at the same time. And right now, they are.

For private market investors: This sets a new benchmark. If Anthropic closes at $900B, every AI company will benchmark against it. OpenAI's $852B suddenly looks cheap. Every Series C AI startup will point to Anthropic when justifying their valuation.

Bottom line: Anthropic going from $380B to $900B in six months proves the AI infrastructure boom is accelerating, not slowing down. Revenue growth is real. Enterprise budgets are massive. And investors are willing to pay any price to be in the room.

Founderscrowd: Ramp, the corporate spend management platform, is in talks to raise $750 million at a pre money valuation of more than $40 billion. The round is being co led by existing backers GIC (Singapore's sovereign wealth fund) and Iconiq Capital.

Six months ago, Ramp was worth $32 billion. Ten months ago, it was worth $16 billion.

The valuation trajectory:

That's 2.5x growth in 10 months. For a corporate card company.

Why this is wild:

Ramp is a fintech that helps businesses manage spending. Corporate cards. Expense reports. Bill payments. Procurement. Treasury management.

Not sexy. Not AI. Not crypto. Just... financial operations software.

And yet, it's growing faster than almost any AI company.

The revenue story:

Ramp crossed $1 billion in annualized revenue in 2025. Customers range from small family farms to space startups (literally). The company has built an all in one platform that combines:

The secret sauce: AI driven automation. Ramp's AI Policy Agent automatically enforces company expense rules in real time. No manual approval chains. No delayed expense reports. Just instant, automated compliance.

Why investors keep paying up:

1. The market is massive.

Corporate card spending in the U.S. alone is a $4 trillion annual market. American Express dominates with legacy infrastructure. Ramp has captured 1.5% to 2% of the market.

Translation: 98% of the market is still untapped.

2. Ramp is profitable (rare for growth startups).

Gross margins hit 64%. The company isn't burning cash to grow. It's printing money while scaling.

3. GIC and Iconiq keep doubling down.

When a sovereign wealth fund (GIC) and an ultra exclusive wealth management firm (Iconiq) keep leading rounds, they're signaling long term conviction. These aren't flip investors. They're IPO prep investors.

The competitive angle:

Ramp's biggest competitor, Brex, got acquired by Capital One for $5.15 billion earlier in 2026. That's an 8x gap between Ramp's valuation and Brex's exit price.

Either Ramp is overvalued or Brex sold too cheap. My bet: Brex sold because they couldn't keep up with Ramp's growth.

Why this matters:

Ramp proves that boring, profitable, AI powered fintech can command AI level valuations. You don't need to be building AGI to hit $40 billion. You just need to solve a real problem, grow fast, and show a path to IPO.

For private market investors: Not every billion dollar valuation is in AI. Fintech infrastructure, especially companies solving enterprise pain points with automation, is still printing money. And if you can get in before the IPO, the returns are real.

Bottom line: Ramp going from $16B to $40B in 10 months shows that enterprise software with real revenue, high margins, and AI powered automation is still the playbook. Boring wins.

Founderscrowd: Kraken parent company Payward just agreed to acquire Hong Kong based stablecoin payments firm Reap Technologies for $600 million in cash and stock. The deal values Payward at $20 billion and marks Kraken's first infrastructure acquisition in Asia.

What Reap does:

Reap provides cross border business payments infrastructure using stablecoins (primarily USDC) as the settlement layer. Think of it as Stripe for stablecoin powered B2B payments.

Product suite:

Reap nearly tripled revenue and transaction volumes in 2025. The global stablecoin card market now exceeds $18 billion annually.

Why Kraken paid $600M:

1. Asia is Kraken's fastest growing market.

Kraken co CEO Arjun Sethi told Bloomberg: "If you take Europe out, the fastest growing market is Asia, not just revenue but also asset on platform."

Reap gives Kraken instant access to Hong Kong, Singapore, Mexico, and emerging markets across Asia, Latin America, and Africa.

2. Stablecoins are becoming payment rails.

Stablecoins (USDC, USDT) are no longer just crypto trading tools. They're becoming the infrastructure layer for global B2B payments.

Why? Because they settle instantly, cost less than traditional wire transfers, and work across borders without bank intermediaries.

3. Kraken is preparing for IPO.

Sethi said Kraken is "about 80% ready" to go public. The Reap acquisition is part of a broader strategy to build a full stack financial infrastructure platform before listing.

The acquisition spree:

Reap is Kraken's fourth major acquisition in about a year:

Total spend: $2.6B+

Payward is building an end to end crypto native financial services platform: trading, custody, derivatives, payments, tokenized assets, and now stablecoin powered B2B infrastructure.

The founders:

Reap was founded by Daren Guo (former Stripe Asia Pacific lead) and Kevin Kang (ex investment banker). Guo built Stripe's APAC business from the ground up. He knows payments. He knows Asia. And now he's bringing that expertise to stablecoins.

Reap will continue operating as a standalone platform under Guo's leadership, but with access to Kraken's global liquidity, custody, and regulatory licenses.

Why this matters:

The Reap acquisition signals that stablecoin infrastructure is no longer speculative. It's operational. Real businesses are using stablecoins to move billions of dollars across borders every month.

For private market investors: Stablecoin infrastructure (payments, treasury, settlement) is becoming a massive category. Companies like Circle (USDC issuer), Bridge (stablecoin payments), and now Reap (via Kraken) are building the plumbing for a financial system that settles on chain.

Bottom line: When a $20 billion crypto exchange pays $600 million for a stablecoin payments company, stablecoins have officially moved from "crypto thing" to "financial infrastructure." The game is changing.

Friday's Premium deal memo:

A Series D enterprise software company doing $180M ARR, 85% gross margins, preparing for 2027 IPO. Already profitable (rare). Raising final private round at $2.8B pre-money. Minimum investment: $100K. Existing investors: Andreessen Horowitz, Sequoia, Salesforce Ventures.

Why it matters: This is one of the last "sleep well at night" enterprise SaaS deals before IPO. Profitable, growing 60% YoY, customers include 40% of Fortune 500. When enterprise software IPOs come back (they will), this exits at $5B+.

This week, Premium members also got:

Next week:

A biotech company with FDA breakthrough designation. Phase 2 trials showing 73% efficacy. Raising Series C before Phase 3. High-risk, high-reward. If Phase 3 works, this is a 10x. If it fails, it's zero.

$40/month. Lock in before price increases to $100/month in May.

Upgrade to Premium →

Founderscrowd: A four-month-old AI startup with 20 employees and no product just raised $500 million at a $4 billion valuation. The round was led by Google's venture arm GV with backing from Nvidia.

Let that sink in. Four months. 20 people. No product. $4 billion valuation.

The company:

Recursive Superintelligence was incorporated on December 31, 2025. The goal: build AI systems that continuously improve themselves without human intervention.

Not AI that generates text or code. Not AI that analyzes data. AI that builds better AI, autonomously.

The technical term is "recursive self-improvement." The sci-fi term is "the singularity." The venture capital term is "$500 million, please."

The founders:

Dream team of ex-DeepMind and ex-OpenAI researchers:

This isn't a first time founder raising on a pitch deck. These are people who've shipped real AI products at OpenAI, DeepMind, and Salesforce.

What they're building:

Recursive wants to automate the entire AI development pipeline:

Today, frontier AI labs (OpenAI, Anthropic, DeepMind) employ hundreds of researchers to manually tune models, curate datasets, run experiments, and decide what to build next.

Recursive's bet: automate all of it. Let the AI figure out how to improve itself.

If it works, you don't need hundreds of researchers. You need 20 people and a lot of computing.

Why GV and Nvidia paid $4B:

1. Pedigree.

When Richard Socher and Tim Rocktäschel walk into your office and say, "We're building self-improving AI," you don't ask for a demo. You write the check.

2. The upside is infinite.

If recursive self-improvement actually works, whoever gets there first controls the entire AI market. That's not a $4 billion outcome. That's a $4 trillion outcome.

3. The round was oversubscribed.

The $500M round was so oversubscribed it could reach $1 billion. Translation: Investors were fighting to get in.

The skeptical take:

Critics warn that this is the AI bubble peak. A company with no product, no revenue, and no public demonstrations just raised half a billion dollars based entirely on the resumes of its founders and the promise of a technology that's never been proven at scale.

Sound familiar? That's because it's exactly what happened in 2021 with crypto and 2024 with AI wrappers.

The difference: Recursive's founders actually know what they're doing. Socher built real products at Salesforce. Rocktäschel's team won the ICML 2024 Best Paper Award. These aren't grifters. They're legit researchers.

But being legit doesn't mean the idea will work.

The public launch:

Recursive plans a public launch around mid May 2026. We'll see demos, technical details, and actual progress (or lack thereof) soon.

Why this matters:

When a four-month-old lab raises $500M at $4B, it tells you two things:

For private market investors: The AI game is consolidating around a handful of players with massive capital and top tier talent. If you're not in OpenAI, Anthropic, Google, or one of the new labs (Recursive, AMI Labs, Ineffable Intelligence), you're probably not going to be relevant in 2027.

Bottom line: Recursive raising $500M at $4B four months after founding is either visionary or insane. We'll know which in 12 months. But the fact that Google and Nvidia made the bet tells you where the smart money thinks AI is heading: autonomous, recursive, self improving systems.

Anthropic's board decision:

Ramp's IPO timeline:

Kraken's IPO prep:

Recursive's public launch: