Anthropic filed its confidential S-1 on Monday, and in doing so opened the most consequential valuation argument the public markets have been handed since the internet first went institutional. The company raised $65 billion in its Series H at a $965 billion valuation days before the filing—eclipsing OpenAI's $852 billion mark from March—and now sits one SEC review away from becoming the first major generative-AI name to trade on an exchange. The desk has seen rich prints before, but never three potential trillion-dollar listings stacked in a single quarter, and never with this much capital already sunk before a share hits the tape.
Anthropic is running at a $47 billion annualized revenue rate, up from roughly $10 billion a year ago. That is real acceleration, the kind that makes bankers willing to whisper about crossing ten figures on debut. The company is on pace for its first profitable quarter, and has told investors the run rate will surpass $50 billion by the end of next month. Growth this steep in a category this large is what turns venture exits into index events—and what makes the repricing risk on the other side of lockup expiry a live grenade.
The race itself tells you everything about positioning. Anthropic filed ahead of OpenAI, claiming first-mover advantage in a market that will compare them in real time and likely price them against each other for years. SpaceX is targeting June 12 at a $1.75 trillion valuation and a $75 billion raise; Anthropic and OpenAI are both eyeing the fall. The concentration is unprecedented—PitchBook calls it "the largest concentration of pre-IPO capital ever brought to market simultaneously"—and the three are likely to raise more collectively than all IPOs did in 2021, the record year.
The read from here: Anthropic gets the pop, then it gets the pain. Hyped tech IPOs surge immediately following their debuts, fueled by scarcity, media frenzy, and FOMO-driven retail participation. Once lockup periods expire, insiders and early venture investors cash out, flooding the market just as initial narratives meet operational realities. The smarter approach is patience—waiting for post-IPO hype to dissipate so the company can be judged on business fundamentals rather than valuation theater. History is consistent on this: Palantir, Snowflake, and Cerebras all gave back the Day One euphoria once the lockup broke and quarterly scrutiny began. The pattern doesn't care how good the model is.
The broader stakes are index-level. The listed AI complex is likely to rerate on Anthropic comps, not just the company itself. Amazon and Alphabet are both investors and Claude distribution partners; they will trade on this whether the correlation makes sense or not. PitchBook warns that 2026 either becomes the most consequential IPO cycle since the dot-com era or the most expensive lesson in narrative-versus-fundamentals that public markets have ever taught. The line between those outcomes sits exactly where it always does: whether the revenue growth these companies are printing today can compound at anything close to the rate their valuations are pricing in.
- The lockup trade: The 180-day lockup expiry often offers a calmer entry for long-term positions—expect the flood six months post-pricing, not on Day One.
- Indirect exposure now: Amazon hosts Claude on Bedrock and is a major investor; Alphabet competes through Gemini and holds a stake—both give you the theme without the IPO lottery.
- The profitability question: Anthropic, OpenAI, and xAI are all spending heavily and have yet to demonstrate sustained profitability, the core fuel for bubble concerns once quarterly disclosure begins.
Global IPOs raised $42.6 billion in Q1 2026, up 45% year-over-year despite a 15% drop in deal count—investors are backing fewer but much larger listings. Anthropic is the logical extreme of that trend: the richest exit in the richest category in the richest window anyone alive has traded through. The valuation will either prove visionary or become the high-water mark everyone uses to time the turn. Either way, it prints first—and that's the edge they came to take.