Load OpenVINO data to DuckDB

OpenVINO Model Server is a high-performance system for deploying AI models that provides RESTful APIs for inference and metadata management, including support for TensorFlow Serving, KServe, and OpenAI-compatible endpoints. Everything needed to build a working OpenVINO → DuckDB pipeline is on this page: the API's base URL, authentication, endpoints, pagination and incremental field — plus a prompt that hands the whole job to your coding agent.

---

Build your OpenVINO to DuckDB pipeline

Paste this prompt into Claude, Codex, or Cursor. The agent does the rest.

PromptCopy prompt
Run uvx dlthub-init@latest to build a pipeline from OpenVINO to DuckDB and run it on dltHub

That scaffolds a dltHub workspace and installs the dltHub AI harness — the project rules, the secrets-management skill, and the dlt MCP server your agent needs to work safely. From there it reads the OpenVINO API, proposes the endpoints to load, then writes, runs and validates the pipeline while you review rather than type. Credentials are inspected through MCP tools, so your agent never reads secrets.toml itself. How the LLM-native workflow works →

Prefer to write it yourself? Every fact the agent uses is below.

---

OpenVINO API at a glance

These values come from the OpenVINO API reference — the authoritative source if anything here looks out of date.

---

How do I authenticate with the OpenVINO API?

Authorization is supported for generative endpoints starting with /v3 via an API key, which must be provided in the 'Authorization' header using the 'Bearer <api_key>' format. The API key is configured server-side via the --api_key_file parameter or the API_KEY environment variable.

1. Get your credentials

OpenVINO Model Server (OVMS) does not have a centralized cloud-based dashboard for API key management. Authentication is enforced primarily for Generative AI (v3/OpenAI-compatible) endpoints. To set up credentials, you must provide an API key directly to the server instance via one of two methods: 1) Save the API key in a text file and provide the path to the file using the --api_key_file CLI argument when starting the server. 2) Set the API key as an environment variable named API_KEY in the environment where the server runs. When calling protected endpoints, the client must include the key in the request header as Authorization: Bearer <api_key>. If no API key is configured, authorization is not enforced and endpoints remain accessible.

2. Add them to .dlt/secrets.toml

Copy
[sources.openvino_source]
api_key = "your_actual_api_key_here"

dlt reads this file automatically at runtime. With the harness, the setup-secrets skill prompts you for the values and never handles the raw credential in chat. For production, see setting up credentials with dlt.

---

What OpenVINO data can I load into DuckDB?

These are the OpenVINO endpoints dlt can load into DuckDB:

---

How do I load only new OpenVINO records?

The OpenVINO API reference does not document a timestamp or sequence field for these endpoints, so there is nothing to advertise here as verified. Pick a field from the endpoints table above that increases with every write, then set it as the cursor_path.

Copy
{"name": "models_list", "endpoint": {
    "path": "v3/models",
    # Replace with a field that increases on every write.
    "incremental": {"cursor_path": "REPLACE_ME", "initial_value": "2024-01-01T00:00:00Z"},
}}

On the first run dlt loads everything from initial_value; on every run after that it requests only what changed and appends with write_disposition="merge" if you set a primary key. See incremental loading.

---

What does the generated OpenVINO pipeline look like?

A standard dlt REST API pipeline — the same code you would write by hand, loading /v2/models/${MODEL_NAME}/infer (KServe Inference) and /v1/models/${MODEL_NAME}/versions/${MODEL_VERSION}:predict (TFS Predict) from the OpenVINO API into DuckDB:

Copy
import dlt
from dlt.sources.rest_api import RESTAPIConfig, rest_api_resources

@dlt.source
def openvino_source(api_key=dlt.secrets.value):
    config: RESTAPIConfig = {
        "client": {
            "base_url": "http://${REST_URL}:${REST_PORT}/v1 or http://${REST_URL}:${REST_PORT}/v2",
            "auth": {"type": "bearer", "token": api_key},
        },
        "resources": [
            {"name": "models_list", "endpoint": {"path": "v3/models"}},
            {"name": "model_metadata_kfs", "endpoint": {"path": "v2/models/{name}"}}
        ],
    }
    yield from rest_api_resources(config)


def load_openvino_to_duckdb() -> None:
    pipeline = dlt.pipeline(
        pipeline_name="openvino_pipeline",
        destination="duckdb",
        dataset_name="openvino_data",
    )
    load_info = pipeline.run(openvino_source())
    print(load_info)


if __name__ == "__main__":
    load_openvino_to_duckdb()

Run it with python openvino_pipeline.py. The agent iterates on this until it loads cleanly — you review and approve, rather than write it from scratch.

---

How do I query OpenVINO data in DuckDB?

dlt creates one table per resource. Query the loaded data with Python or SQL — or ask your agent to, through the MCP server's execute_sql_query tool.

Python (pandas DataFrame):

Copy
import dlt

data = dlt.pipeline("openvino_pipeline").dataset()
df = data.models_list.df()
print(df.head())

SQL:

Copy
SELECT * FROM openvino_data.models_list LIMIT 10;

See querying your data with dataset and exploring it in marimo notebooks.

---

How do I deploy the OpenVINO to DuckDB pipeline in production?

The pipeline runs locally, which is ideal for prototyping and one-off analysis. When you need it on a schedule, monitored on every load, and shared with your team, deploy the same dlt code on the dltHub platform — no infrastructure to maintain. The prompt above already ends with "run it on dltHub", so your agent can take it there directly.

• Deploy & schedule — run the pipeline as a managed job with automatic retries.
• Monitor — observable job queues, alerting, and load metrics for every run.
• Transform — promote raw OpenVINO loads into governed, documented models.
• Visualize & share — explore data in notebooks and publish live dashboards instead of static screenshots.

Book a demo →

---

What other destinations can I load OpenVINO data to?

dlt loads into any of these — only the destination argument changes:

Set dlt.pipeline(destination="snowflake") and add credentials in .dlt/secrets.toml. On the dltHub platform the same pipeline runs against a managed Iceberg lakehouse. See the full destinations list.

---

Next steps

• Ready for production? Book a demo of the dltHub platform →
• dlt LLM-native workflow docs
• REST API source reference
• Join the dlt community on Slack