Switching Models

Switch Cloudflare Workers AI models and gateway routes per agent.

Code

Pick one slash-separated route per client via id. Switching models means a new Cloudflare(id=...) or a new Agent(model="cloudflare:...") string. For fallbacks, configure an AI Gateway Dynamic Route and pass id="dynamic/<route>".

switch_model.py

from agno.agent import Agent
from agno.models.cloudflare import DEFAULT_GATEWAY_MODEL, Cloudflare

# Second Workers AI model: paste the same string as the catalog "copy" box on the model page.
ALT_WORKERS_MODEL = "@cf/google/gemma-4-26b-a4b-it"

if __name__ == "__main__":
    # Default Workers AI model (no explicit id).
    Agent(model=Cloudflare(), markdown=True).print_response("Say hello in five words.")

    # Pass the gateway route string as `id`.
    Agent(model=Cloudflare(id=DEFAULT_GATEWAY_MODEL), markdown=True).print_response(
        "Say hello in five words."
    )

    # Switch model: new client instance with a different `id`.
    Agent(model=Cloudflare(id=ALT_WORKERS_MODEL), markdown=True).print_response(
        "Say hello in five words."
    )

    # String syntax: everything after the first colon is the gateway `model` id.
    Agent(
        model=f"cloudflare:{ALT_WORKERS_MODEL}",
        markdown=True,
    ).print_response("Say hello in five words.")

    # Dynamic route configured in the Cloudflare dashboard (example id only).
    # Agent(model=Cloudflare(id="dynamic/my-route"), markdown=True).print_response("...")

Usage

1. Set up your virtual environment

   Mac

   uv venv --python 3.12
   source .venv/bin/activate
   

   uv venv --python 3.12
   .venv\Scripts\activate
   
2. Set your environment variables

   export CLOUDFLARE_API_TOKEN=xxx
   export CLOUDFLARE_ACCOUNT_ID=xxx
   export CLOUDFLARE_AI_GATEWAY_ID=xxx  # optional, defaults to "default"
   
3. Install dependencies

   uv pip install -U openai agno
   
4. Run Agent

   Save the code above as switch_model.py, then run:

   python switch_model.py