When AI Models Proliferate, the Most Valuable Asset Is No Longer the Model Itself
Deep thoughts on AI and aspirations —— ByteDance Deep Thinking Circle
A few years ago, the core question in AI was “whose model is the strongest.” Now that question is quietly being replaced by another: when models are proliferating and their capabilities are converging, how should users choose? Which one? How?
A telling move: a major payments company spent a valuation exceeding ten billion dollars to acquire a “model routing” company. This company doesn’t train models, doesn’t sell models—what it does is more straightforward: it helps users decide “which model should handle this task” among dozens of options. A company that produces no models is worth more than many that do.
This isn’t an anomaly—it’s a structural inflection point in the AI industry: As models converge, value begins migrating to the layer “between models.”
The More Similar the Models, the More Valuable the Choice
When all models can chat, write code, and summarize documents, “using a good model” is no longer an advantage but infrastructure. At that point, the real cost isn’t in the model itself—it’s in selection and switching.
Users face a pile of models: some cheap, some fast, some good at long texts, others at writing code. Choose wrong and you waste money, waste quality, or have to redo the work. These “costs of choice” were previously unaddressed. Then a new role emerged—the layer that makes decisions between models for you.
The value of this layer isn’t in its own strength, but in what it possesses that others don’t: who used which model for what task, how it performed, and what it cost. This cross-model data is unavailable to individual model providers, but the routing layer has it all. With data in hand, it can make more accurate recommendations—that’s the moat.
The Routing Layer’s Real Moat: Not Technology, but Network
A routing tool’s value lies not in algorithmic precision, but in the more models and users it connects, the richer its data, the more accurate its recommendations, which then attracts more users—a positive feedback loop. This is classic network effects.
This also means routing is a “winner-takes-most” business: once one player accumulates sufficiently deep data, its recommendations remain consistently superior, making it hard for latecomers to catch up. That’s why model providers, cloud giants, and independent routing companies are all crowding this track.
But This Layer Has an Underestimated Fracture
The routing layer looks attractive, but it has a fatal weakness: it depends on model providers.
If model providers build their own routing one day (they have every incentive and capability to do so), or if users start bypassing routing to call models directly, this layer’s value evaporates. Its position is somewhat like a “tollbooth”—it makes money as long as cars use the highway, but if someone builds a new road, it becomes obsolete.
So the routing layer’s real test isn’t “can it make good choices,” but “can it make itself indispensable to both model providers and users.” The former relies on data, the latter on whether it’s embedded in users’ workflows.
Judgment for Entrepreneurs
If you’re evaluating this space, my assessment is: the routing layer is a real opportunity, but not an easy business.
Its value lies in data, so you need to connect enough models and users early to capture data accumulation; its risk lies in dependency, so you need to figure out how to avoid being bypassed—embed in workflows, accumulate proprietary data, or become the irreplaceable link in a specific vertical scenario.
Simply building a “model price comparison tool” is indefensible because comparison has no barriers. The real opportunity lies in “understanding better than model providers what users need in which scenarios,” and turning that understanding into something others can’t take away.
Key points: Value migrates to the routing layer as models converge; the routing layer’s moat is cross-model data + network effects; but dependency on model providers is a fracture; opportunity lies in capturing data and embedding in workflows, not in doing price comparison.