Cerebras

Cerebras Inference serves open models at high speed. Behind Verdictan, the Cerebras key stays on the gateway host, and your application calls the gateway with its application token. Send one small chat request before you add Cerebras to your application.

This adapter handles text chat only.

Before you start

• A working gateway from the quickstart.
• A Cerebras API key, and one model ID that your account can use.
• The gateway startup command, and a separate application token for test requests.

Configure the provider

policy-config.yaml

pack:
  name: cerebras-integration
  version: 1.0.0
  enabled: true
policies:
  chain:
    - prompt-injection
providers:
  targets:
    - id: cerebras-primary
      provider: cerebras
      model: "replace-with-cerebras-model-id"
      base_url: https://api.cerebras.ai
      secret_key_ref:
        env: VERDICTAN_CEREBRAS_API_KEY

base_url must start with https://. The gateway appends /v1/chat/completions and sends the key from VERDICTAN_CEREBRAS_API_KEY as a Bearer token.

Start and verify

The Console procedure below also has a CLI 0.1.14 deployment option. Wait for In sync, then send the Cerebras request. A matching configuration does not prove provider access.

With the quickstart in Docker, follow the container steps instead of steps 1 to 3.

1. In the gateway terminal, set the key and lint the file:

   export VERDICTAN_CEREBRAS_API_KEY="<provider-api-key>"
   verdictan policy lint --file policy-config.yaml

2. Restart the test gateway with the key in its environment.

3. Deploy the file to the test agent in Configurations. If that agent already runs pack.version 1.0.0, increase the version first. Then wait for the gateway log line connected read model refresh applied updated config with your version as config_version. Try the change on a test gateway has the steps.

4. In a second terminal, send a request for the configured model:

   export VERDICTAN_CLIENT_TOKEN="<application-token>"
   curl --fail-with-body http://127.0.0.1:41002/v1/chat/completions \
     -H "Authorization: Bearer ${VERDICTAN_CLIENT_TOKEN}" \
     -H "Content-Type: application/json" \
     -d '{"model":"replace-with-cerebras-model-id","messages":[{"role":"user","content":"Reply with Hello."}]}'

choices[0].message.content contains the reply, and the request appears in History when capture is on, or in the delivered events.

Supported requests

Request	Support
Chat Completions	Yes. Streaming and tools work when the model supports them.
Messages	Translated to Chat Completions
Responses, embeddings, image input, response_format, parallel tool calls	No

The Cerebras API can offer more than this adapter. Do not infer gateway support from the Cerebras documentation.

Failure behavior

Problem	What to check
401 from the gateway	The application token is active, and the key is set on the gateway host.
Provider authentication error	VERDICTAN_CEREBRAS_API_KEY is set in the gateway process. Restart the gateway.
Model not found	The request model matches the target, and your account can use it.
capability.unsupported_family	The request is Chat Completions or Messages.
Provider 429 or another upstream error	The response body and your Cerebras rate limits. See Error Handling.

Limits

• Model names, prices, rate limits, and data terms are in the Cerebras Inference documentation.
• Streaming and tools vary by model. Test each one with your model.
• A listener on a non-loopback address needs TLS and network controls.

Next steps

• Provider Configuration: target fields, routing, and fallback.
• Runtime Request Families: request families and their policy stages.