Start with one small request.

Data access requires an invited account with active access to an admitted dataset. If your account has no available data, review its status or contact support@originfold.com.

1. Open your invitation and sign in with its verified email. Complete the account steps shown.
2. Open the Data Explorer. Choose a dataset, enter exact entity IDs and start with 25 rows over a short date range.
3. Inspect the result's version, source context and availability time. Loading another page makes one more request.
4. For a notebook or application, create a named key in Your account. Store it in an environment variable when its secret is shown.

curl "$ORIGINFOLD_API_URL/v1/data-catalog" \
  --header "Authorization: Bearer $ORIGINFOLD_API_KEY"

Use your environment's API origin for ORIGINFOLD_API_URL. The catalogue gives the exact dataset/version, field types and coverage you may query. The API origin is the service address supplied for your environment; do not assume the marketing website is the API. Do not type your actual key into these examples or save it in shell history. After a query, the Explorer can copy its complete HTTP equivalent without exposing your key.

Use Python 3.12 or newer. From a source checkout, install the SDK in your terminal with python -m pip install ./packages/originfold_sdk. Then run this Python code:

from originfold_sdk import Client

with Client.from_env() as client:
    print(client.catalogue())
    print(client.usage())

The Python SDK is included in the source checkout and is not yet published to PyPI. It reads ORIGINFOLD_API_KEY and ORIGINFOLD_API_URL from your environment. It uses explicit typed queries and pages, with no hidden retries or automatic downloads. Returned decimal numbers retain their precision.

Never put a key in a shared notebook, URL, commit or support message. Rotate it from your account; the old key has at most 24 hours of overlap. Revoke it promptly if exposed.