Platform WatchJune 28, 2026via The Decoder
Coinbase joins the rush to Chinese AI models as Western labs face a pricing stress test
Why it matters
Enterprise AI cost optimization is forcing a strategic pivot away from Western closed models. Coinbase's automated routing system and caching infrastructure show how scale operators are arbitraging model pricing—and it's working at the token level, not just the contract level.
Key signals
- Coinbase switched to Chinese AI models (GLM 5.2, Kimi 2.7)
- Implemented automated routing system for model selection based on task and price
- Cache hit rate improved from 5% to 60%
- Achieved 50% reduction in total AI spending while token usage increased
- CEO Brian Armstrong leading the initiative
The hook
Coinbase just cut its AI spending in half. Here's how: switching to Chinese models and a 12x cache hit rate improvement.
Coinbase CEO Brian Armstrong is switching his company to Chinese AI models like GLM 5.2 and Kimi 2.7. An automated routing system picks the best model for each request based on task and price, and better caching pushed the hit rate from 5 to 60 percent. Coinbase has cut its AI spending in half even as token usage keeps climbing.