Summary
Cloudflare merged Workers AI and AI Gateway into one AI control plane with a shared binding and REST endpoint. A default gateway is now created automatically for observability, AI Gateway credits pay for both Cloudflare-hosted and external models, and unified billing unlocks higher rate limits, with model-first and smart routing coming next.
What changed
On August 7, 2026 Cloudflare unified Workers AI and AI Gateway: they now share one binding and REST API, a default gateway auto-generates on the first authenticated request (adding request logging, token tracking, and cost attribution with no setup), AI Gateway credits apply to Workers AI usage alongside external providers, and unified billing raises Workers AI rate limits. Model-first routing and a prompt-classifier smart router were previewed.
Why it matters
Collapses the choice between Cloudflare's own inference and third-party providers into one wallet, one API, and one observability plane, cutting integration overhead and making Cloudflare a neutral routing layer across OpenAI, Anthropic, and open models.
Evidence excerpt
You can request Kimi K2.7 Code and not care whether it comes from Workers AI, another provider.