Why the rise of open source AI isn’t hurting Anthropic … yet
TL;DR
Decagon CEO Jesse Zhang frames enterprise AI as a lifecycle: expensive frontier models prove new use cases, then cheaper open source models take over mature production workloads. Vercel’s AI gateway shows the volume shift: DeepSeek now handles more than a third of tokens on the platform, while Z.ai jumped to fourth place in a week. Spend tells a different story: Anthropic still accounts for more than half of AI spending on Vercel, even after a small recent share drop.
Nauti's Take
The piece lands on a useful market pattern, but Zhang’s argument is still light on hard evidence. It matches what many teams see in practice: the best model often acts as the scout, while the cheaper model later runs the repeatable work in the background.
Anthropic is not under real pressure yet as long as new, difficult use cases keep arriving faster than old ones become commodity workloads. The pressure starts when open source does more than cut production costs and begins discovering new workflows too.
Briefingshow
For AI users, model choice is becoming less ideological and more workflow-dependent. Teams can discover new processes with Claude, Opus, or other top models, then move parts of the workload to cheaper open source models. The key question is which tasks are stable enough to migrate without breaking quality.
Sources