API Reference

Creating Datasets

Version: 4.1
Updated: 17th August 2026

Overview

Post a partial dataset JSON to create a new dataset.

The only required field is title.
All other fields are optional.
- id is automatically assigned.
- sharing defaults to public.
- topics not defined in the site settings are silently dropped.
You need permission to create datasets: membership of a group with canCreateDatasets. See Permissions.
- If you are not a dataset manager, you are automatically granted the editor role on the new dataset.

Note

You cannot upload files while creating the dataset.
First, the dataset must be assigned an ID. Then you can use the upload files API.

POST

https://data.website/api/v3/dataset

HTTP Headers

Authorization	API key.	string	Required
Content-Type	Must be "application/json". Most clients will set this automatically.	string	Required

HTTP Return Codes

200	Returns the full dataset JSON.
400	Returns a JSON object explaining the error.
403	API key does not have permission to create datasets.

Example: Minimal dataset

import requests, os, json

api_key = os.getenv("DATAPRESS_API_KEY")

headers = { "Authorization": api_key }
dataset = {
    "title": "My new dataset",
    # Optional:
    "logo": { "title": "My Organisation", "img": None },
    "topics": ["waste-and-recycling"],  # from /api/v3/site
}

url = f"https://data.website/api/v3/dataset"

response = requests.post(url, json=dataset, headers=headers)
print(json.dumps(response.json(), indent=2))

{
  "dataset": {
    "id": "444xy", // Automatically assigned
    "title": "My new dataset",
    "logo": { "title": "My Organisation", "img": null },
    "topics": ["waste-and-recycling"],
    // [...full JSON structure...]
  }
}