# Cloudflare AI Gateway (basic) (/examples/models/cloudflare/basic)



Default model is **Workers AI** (only Cloudflare token + account id). You can paste a catalog binding: `Agent(model="cloudflare:@cf/meta/llama-3.3-70b-instruct-fp8-fast")`, the full gateway id, or `Cloudflare()` defaults.

<Note>
  The current Agno `Cloudflare` adapter uses AI Gateway's `/compat/chat/completions` endpoint. Cloudflare [deprecates this endpoint for single-model calls](https://developers.cloudflare.com/ai-gateway/usage/chat-completion/) while keeping it working for existing integrations. Dynamic routes still require it. For new direct model integrations, follow Cloudflare's [current REST API guide](https://developers.cloudflare.com/ai-gateway/usage/rest-api/).
</Note>

```python title="basic.py"
"""
Cloudflare AI Gateway (basic)
=============================

Cookbook example for Cloudflare AI Gateway OpenAI-compatible unified API.

Requires:
- CLOUDFLARE_API_TOKEN
- CLOUDFLARE_ACCOUNT_ID

Optional:
- CLOUDFLARE_AI_GATEWAY_ID (defaults to the ``default`` gateway)

Default model is **Workers AI** (only Cloudflare token + account id). You can paste a catalog binding:
``Agent(model="cloudflare:@cf/meta/llama-3.3-70b-instruct-fp8-fast")``, the full gateway id, or ``Cloudflare()`` defaults.

Set ``CLOUDFLARE_API_TOKEN`` and ``CLOUDFLARE_ACCOUNT_ID`` in your shell before running (see README),
not in this file.

For switching models (OpenRouter-style ``id`` string), see ``switch_model.py`` in this folder.

For ``google/...`` or ``openai/...`` you need a supported gateway provider name, BYOK keys in the
dashboard, and a model id from the Cloudflare docs — otherwise you may see **Invalid provider** (HTTP 400)
or upstream **401** errors.
"""

import asyncio

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

# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------

agent = Agent(
    model=Cloudflare("@cf/meta/llama-3.3-70b-instruct-fp8-fast"),
    markdown=True,
)

# ---------------------------------------------------------------------------
# Run Agent
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    # --- Sync ---
    agent.print_response("Share a 2 sentence horror story")

    # --- Sync + Streaming ---
    agent.print_response("Share a 2 sentence horror story", stream=True)

    # --- Async ---
    asyncio.run(agent.aprint_response("Share a 2 sentence horror story"))

    # --- Async + Streaming ---
    asyncio.run(agent.aprint_response("Share a 2 sentence horror story", stream=True))
```

## Run the Example [#run-the-example]

<Steps>
    <Step title="Set up your virtual environment">
      <CodeBlockTabs defaultValue="Mac">
        <CodeBlockTabsList>
          <CodeBlockTabsTrigger value="Mac">
            Mac
          </CodeBlockTabsTrigger>

          <CodeBlockTabsTrigger value="Windows">
            Windows
          </CodeBlockTabsTrigger>
        </CodeBlockTabsList>

        <CodeBlockTab value="Mac">
          ```bash
          uv venv --python 3.12
          source .venv/bin/activate
          ```
        </CodeBlockTab>

        <CodeBlockTab value="Windows">
          ```bash
          uv venv --python 3.12
          .venv\Scripts\activate
          ```
        </CodeBlockTab>
      </CodeBlockTabs>
    </Step>

  <Step title="Install dependencies">
    ```bash
    uv pip install -U agno openai
    ```
  </Step>

    <Step title="Set up Cloudflare access">
      Use a Cloudflare account and an API token with the permissions required for [AI Gateway and Workers AI](https://developers.cloudflare.com/ai-gateway/usage/chat-completion/). The default gateway is created on the first request; set `CLOUDFLARE_AI_GATEWAY_ID` only to use another existing gateway.

      <CodeBlockTabs defaultValue="Mac/Linux">
        <CodeBlockTabsList>
          <CodeBlockTabsTrigger value="Mac/Linux">
            Mac/Linux
          </CodeBlockTabsTrigger>

          <CodeBlockTabsTrigger value="Windows">
            Windows
          </CodeBlockTabsTrigger>
        </CodeBlockTabsList>

        <CodeBlockTab value="Mac/Linux">
          ```bash
          export CLOUDFLARE_ACCOUNT_ID="your_cloudflare_account_id_here"
          export CLOUDFLARE_API_TOKEN="your_cloudflare_api_token_here"
          ```
        </CodeBlockTab>

        <CodeBlockTab value="Windows">
          ```powershell
          $Env:CLOUDFLARE_ACCOUNT_ID="your_cloudflare_account_id_here"
          $Env:CLOUDFLARE_API_TOKEN="your_cloudflare_api_token_here"
          ```
        </CodeBlockTab>
      </CodeBlockTabs>

      Workers AI examples use this Cloudflare token. Other providers can require stored provider keys or supported unified billing; configure the selected route in the gateway before switching to it.
    </Step>

  <Step title="Run the example">
    Save the code above as `basic.py`, then run:

    ```bash
    python basic.py
    ```
  </Step>
</Steps>

Full source: [cookbook/90\_models/cloudflare/basic.py](https://github.com/agno-agi/agno/blob/8f36eaf2d18e91afa7b327eec66a3cd3685dcb87/cookbook/90_models/cloudflare/basic.py)