Coinbase switches default AI model to open-source models such as GLM5.2 and Kimi2.7, halving AI costs

robot
Abstract generation in progress
Mars Finance News, June 28 - Coinbase CEO Brian Armstrong disclosed that the company has reduced AI spending by nearly half despite continuous exponential growth in token usage, by switching the default AI model to open-source models such as GLM 5.2 and Kimi 2.7, combined with intelligent model routing and cache optimization. Brian Armstrong stated that Coinbase no longer relies on usage limits or consumption alerts to control costs; instead, it defaults to lower-cost open-source models and only calls on more powerful frontier models in necessary scenarios. At the same time, the company further reduces inference costs through automatic model routing, improving cache hit rates, and streamlining context. Notably, this statement comes amid recent public expressions of concern from the CEO of Anthropic about open-source models and model distillation. Coinbase’s practice shows that in enterprise AI applications, open-source models are becoming an important choice for balancing cost and efficiency.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments