OpenAI GPT-6 Luna is now available on Foundation Model APIs

Use cases

Imagine you need a fast, cost-efficient model for reasoning and agentic tasks without paying the high cost of flagship models. GPT-6 Luna provides:

• Fast reasoning and agentic capabilities at the lowest cost in the GPT-6 lineup.
• Support for multimodal inputs (text and image).
• Access via pay-per-token endpoints in your workspace.

How to use

You can query the model using its system service name in your API requests.

from databricks.sdk import WorkspaceClient

w = WorkspaceClient()

response = w.serving_endpoints.query(
    name="databricks-gpt-6-luna",
    messages=[
        {
            "role": "user",
            "content": "Explain this diagram in one sentence."
        }
    ]
)
print(response)

What happens in this code:

1. Initializes the Databricks workspace client.
2. Sends a query request to the GPT-6 Luna model endpoint.
3. Passes a user text message containing the prompt.
4. Prints the response generated by the model.

Limitations and requirements

• This model is not supported in AI Playground; use the Responses API to interact with it.
• Customers are responsible for ensuring compliance with applicable model terms.
• Available in pay-per-token mode.