Why open source AI isn’t hurting Anthropic’s revenue yet

A curious tension is emerging in the AI economy: mature enterprise deployments are steadily migrating toward lighter, cheaper open source models, yet overall spending on frontier models like Anthropic‘s remains stubbornly high.

Decagon CEO Jesse Zhang argues these two tiers aren’t competitors but rather two phases of the same lifecycle — expensive frontier models prove out new use cases, which are then handed off to cheaper alternatives once they stabilize.

Data from Vercel‘s AI gateway supports the pattern: DeepSeek V4 Flash now processes over a third of token volume, whereas Anthropic still commands more than half of total AI spend.

The volume is shifting toward open source, but the revenue is not.