Skip to main content

​
Cloudflare AI Gateway

Cloudflare AI Gateway sits in front of provider APIs and lets you add analytics, caching, rate limits, and controls. Fased’s current built-in setup uses the Anthropic Messages API through your Gateway endpoint. Cloudflare AI Gateway is a current Fased model provider route. The built-in route is intentionally narrow: it configures Cloudflare Gateway as an Anthropic Messages path, then exposes it in Agent > Models.
  • Provider: cloudflare-ai-gateway
  • Base URL: https://gateway.ai.cloudflare.com/v1/<account_id>/<gateway_id>/anthropic
  • Default model: cloudflare-ai-gateway/claude-sonnet-4-6
  • API key: CLOUDFLARE_AI_GATEWAY_API_KEY (the Anthropic provider key used for requests through this Gateway path)
  • Auth method: Account ID + Gateway ID + API key
Cloudflare also supports OpenAI-compatible Gateway routes, Workers AI routes, and authenticated Gateway tokens. Those are not the default Fased Cloudflare setup yet. Use the explicit models.providers.cloudflare-ai-gateway config if you intentionally need a different Cloudflare Gateway path.

​
Quick start

  1. Set the provider API key and Gateway details. In the browser, use Agents > selected Agent > Models > Cloudflare AI Gateway. For CLI setup:
  1. Open Agents, select the Agent, then use Agent > Models to choose the Cloudflare AI Gateway model for primary, fallback, or task work.
For scripted setup, you can set the Agent default model in config:

​
Non-interactive example

​
Authenticated gateways

If you enabled Gateway authentication in Cloudflare, add the cf-aig-authorization header. This is in addition to the provider API key used by the upstream model provider.

​
Model scope

The normal Fased picker shows the supported built-in Cloudflare model:
  • cloudflare-ai-gateway/claude-sonnet-4-6
Cloudflare Gateway can route to many upstream providers, but Fased does not pretend there is one global Cloudflare model catalog. If you configure a custom Cloudflare Gateway route, add the exact model entries in Advanced > Config so Agent > Models and Chat know which model IDs are valid for your Gateway.

​
Environment note

If the Gateway runs as a daemon (launchd/systemd), make sure CLOUDFLARE_AI_GATEWAY_API_KEY is available to that process. Use ~/.fased/.env or env.shellEnv.