The AI company that burned through billions to build Claude is now asking investors to price it like Apple, based on revenue that doesn't exist yet.
The Summary
- Anthropic is preparing for an October IPO with a target valuation above $2 trillion, which would make it the largest public offering in history
- The valuation hinges on revenue projections of $190-200B by 2028, a massive leap from current run rates that even Jim Cramer says justifies the price
- This represents a fundamental shift in how markets value AI companies: growth trajectory over current profits, with investors betting on future potential rather than present financial metrics
The Signal
Anthropic is asking public markets to do something private investors have been doing for three years: believe that the company building Claude can become bigger than Microsoft, Meta, or Alphabet. The proposed $2 trillion valuation would place Anthropic in rarefied air, competing only with Apple and NVIDIA for the title of most valuable company on Earth. The bet isn't on what Anthropic is today. It's on what the revenue trajectory looks like if Claude becomes the default AI interface for enterprise work.
The revenue forecast of $190-200B by 2028 is doing the heavy lifting here. For context, OpenAI hit a $3.4B annual run rate in late 2024. Anthropic's current revenue is likely in the same ballpark, maybe slightly behind. That means investors are being asked to believe in roughly 50-60x growth in under four years. Not impossible in AI land, but it requires everything to go right: enterprise adoption accelerates, compute costs drop, and Claude maintains competitive edge against OpenAI, Google, and every other foundation model player.
"The market is prioritizing future growth potential over current financial metrics."
What makes this valuation structure interesting for Web4 is what it signals about infrastructure value. Anthropic isn't just selling API calls. The company is betting that rapid revenue growth will come from being the intelligence layer that agents run on. If you believe the agent economy is real, and that every company will eventually deploy dozens or hundreds of AI workers, then the firms that own the core models own the equivalent of AWS in the cloud era. That's a $200B+ revenue business if it plays out.
The timing matters too. October 2026 puts this IPO right as the first wave of production AI agents hits enterprise scale. Companies have moved past pilots. They're buying seat licenses for AI workers that handle customer service, research, data analysis, and code review. Anthropic's pitch is that Claude is the brain those agents need, and that brain should trade at 10x forward revenue just like SaaS companies did in the 2010s.
Key investor dynamics:
- Jim Cramer's endorsement signals mainstream finance is ready to price AI like tech, not like science experiments
- Private market investors who backed Anthropic at lower valuations stand to see 3-5x returns at IPO
- Public market buyers are betting the $2T valuation looks cheap if 2028 revenue actually lands at $200B
The Implication
If Anthropic pulls this off, it resets the valuation bar for every AI infrastructure company. Cohere, Mistral, and the next wave of foundation model startups will point to Anthropic's multiples and say "we deserve the same." That could flood capital into the picks-and-shovels layer of Web4, which is good if you're building agents or tooling. It's bad if you're trying to compete with companies that can burn $10B a year without blinking.
For anyone thinking about where to build in the agent economy, watch how the market reacts to this IPO. If investors buy the $200B revenue story, it means they believe agents are going to eat a massive chunk of knowledge work. If the stock craters post-IPO, it means the market still sees AI as a feature, not a platform. Either outcome tells you where the real money is betting.