Anthropic just told you where it thinks the real money is, and it isn't in the model.
Chris Taylor runs a company called Ode with Anthropic. He isn't shy about the ceiling. "It's pretty easy to imagine this as a trillion-dollar company someday if we execute well," he told TechCrunch. That's a bold line. The firm employs about 100 engineers, as of this month. But Taylor has done this before. He co-founded Fractional AI in 2024, spent eleven months building an AI services shop on top of OpenAI's models, and then watched Anthropic buy the whole operation in May to serve as the founding team of Ode.
Ode launched on May 4, with Anthropic, Blackstone, Hellman & Friedman and Goldman Sachs as its backers, joined by a consortium that includes General Atlantic, Leonard Green, Apollo Global Management, GIC and Sequoia Capital. That's serious backing. The firm carries roughly $1.5 billion in committed capital. Its engineers don't sell software licenses. They show up at mid-size companies, sit with the operations team, and build Claude-based systems tuned to whatever that specific business actually does: invoicing, claims processing, customer service routing, the unglamorous plumbing that runs a company.
That's the bet: the bottleneck in enterprise AI was never Claude's reasoning ability. It's the six months of unglamorous integration work between a capable model and a company's actual invoicing system.
Every Lab Wants a Piece of the Services Layer #
Anthropic isn't alone in reaching this conclusion. The timing makes it look less like a hunch and more like an industry consensus. That's not a coincidence. OpenAI has finalized its own version, the OpenAI Deployment Company, backed by a $10 billion vehicle anchored by TPG, with Advent, Bain Capital and Brookfield as co-lead partners and Goldman Sachs, SoftBank and Warburg Pincus among the founders. OpenAI acquired Tomoro to bring roughly 150 Forward Deployed Engineers into the venture from day one, a title and a delivery model borrowed directly from Palantir's playbook of embedding engineers inside client organizations rather than shipping them a product to configure themselves.
Anthropic and OpenAI have landed in the same place: two private-equity-anchored joint ventures built on one shared thesis. Neither company wants to be just a model vendor selling API calls by the token. Both want a cut of the services layer that used to belong exclusively to firms like Accenture, and both are willing to let buyout shops write the checks to get there.
Goldman Sachs shows up as a backer in both deals. That tells you something. A bank with deep relationships across corporate America is reading this moment a certain way. It isn't picking a side between Anthropic and OpenAI. It's betting on the category itself: implementation as the next platform layer, regardless of which lab's model ends up underneath.
For Accenture, Deloitte and the rest of the traditional systems integrator world, this is the uncomfortable part. Those firms built decades of business on exactly this kind of embedded, high-touch technology rollout. That business is now under threat. Now the labs that make the underlying models are deciding they'd rather keep that margin themselves than hand it to a consulting partner. Ode's private equity backers aren't just writing checks either. Blackstone and Hellman & Friedman own portfolios full of mid-size companies that are precisely the customer Ode is built to serve, which means this is as much a distribution deal as it is a technology bet.
The Numbers Behind the Bet #
Anthropic's own numbers explain why it can afford to make this bet at all. The growth is real. The company has reportedly reached a $30 billion revenue run rate, according to VentureBeat, and closed a $65 billion raise in May that pushed its valuation toward $1 trillion ahead of a planned IPO, as TechCrunch reported. A company growing that fast can afford to gamble $1.5 billion on a services subsidiary that might take years to prove out.
Frankly, the trillion-dollar line from Taylor should be read as marketing bravado dressed up as forecasting. That's just marketing. No 100-person firm knows it's going to be worth a trillion dollars. What it does know is that its parent company needs enterprise customers to actually deploy the technology it's selling, not just license it and let it sit unused. If Ode and the OpenAI Deployment Company succeed, the AI labs stop being toolmakers and start being the contractors too. That's the real wager. The consulting firms that used to own that relationship will have to explain why anyone still needs them in the room.
Also read: Rime Raises $24 Million to Put Its AI Voice on Enterprise Phone Lines • Indian Coding Startup Emergent Becomes a Unicorn in Just Over a Year • Kevin Ryan's AlleyCorp Raises $335 Million and Still Won't Chase Mega-Rounds