Cloudflare AI Gateway
:::info Status: roadmap (stub) The Cloudflare adapter is scaffolding, and scope is limited by what Cloudflare AI Gateway allows. The reference adapter today is LiteLLM. :::
Intended integration (limited)
| Capability | Mechanism (planned) | Notes |
|---|---|---|
| Request transform | — | ❌ Not supported — Cloudflare AI Gateway has no custom routing hook to call POST {OPTIMIZER}/v1/optimize and rewrite the request. |
| Cache hit | managed config | Partial — via Cloudflare's own caching. |
| Response transform | managed config / logs | Limited — Cloudflare's managed surface. |
Because Cloudflare AI Gateway is managed config with no custom-routing hook, Anyray
cannot transform requests (/v1/optimize) there. The integration is limited to
Cloudflare's own caching and observability.
If request-level optimization matters to you on Cloudflare, run a gateway that supports a routing hook (LiteLLM, Kong, Envoy) behind it, or use Anyray's own gateway.
Want this sooner?
See adding a new gateway and track status in
integrations/README.md.