Your AI bill is out of control. Cloudflare can fix it now.
This title could be clearer and more informative.Try out Clickbait Shieldfor free (5 uses left this month).
Cloudflare is launching spend limits in AI Gateway to help companies control runaway AI token costs. The feature lets teams set dollar-based budgets scoped by model, provider, or custom attributes like user or team, with fixed or rolling time windows. When a budget is hit, requests can be blocked or routed to a cheaper fallback model via Dynamic Routes. A closed beta also introduces identity-driven budgets using Cloudflare Access, which attaches authenticated user identity (from JWT) to every AI request — enabling per-user and per-team cost attribution without custom code. This allows policies like giving ML engineers access to frontier models while routing interns to open-source alternatives. Intelligent task-based routing to automatically select the cheapest suitable model is also in development.
