For AI Agents
A data marketplace
your agent can use.
DataBazaar is built API-first. Your agent can search, preview, purchase, and even list datasets programmatically — via MCP server or plain REST. Search and samples are public. Create an account to retrieve full datasets, including free ones.
Install the DataBazaar skill for Codex or Claude Code, try a dataset workflow, or use our framework integration examples.
Start here · No account or API key
Preview now. Sign in to get the full dataset.
Explore a real CSV of 2026 US federal holidays. This Python 3 example searches the catalog and parses its public sample. It prints sample rows and elapsed time, then shows how to connect your account for the complete dataset. No packages to install.
curl -fsS --max-time 20 https://databazaar.io/examples/first-retrieval.py -o first-retrieval.py python3 first-retrieval.py
Expected result: "scope": "sample_only", plus dates and column names. Read the example source or inspect the listing.
If the example fails
A network timeout or HTTP 429 means retry after a minute. If the listing is unavailable, search for another free CSV using GET /datasets?pricing_type=free. A download URL expires after one hour; request a fresh one. No payment is attempted by this example.
Connect your agent directly
Add this Streamable HTTP MCP URL to your agent. Choose no authentication for public search and samples. No local server or npm install is needed.
https://api.databazaar.io/mcp/publicThen give your agent this first task:
Search for US federal holidays 2026 with pricing_type=free. Use dataset 8bc44aae-5b51-4b9d-9d4a-9078f3f9f553. Call preview_sample, fetch its sample_url, and inspect the CSV sample. Report sample rows and columns. To retrieve the full dataset, ask me to connect my account through https://api.databazaar.io/mcp first.
Check the connection from your terminal with python3 first-retrieval.py --mcp. It verifies tool discovery and previews the same sample through MCP.
Ready to use the full dataset?
- Create a free account or sign in.
- Add
https://api.databazaar.io/mcpto your agent and complete the OAuth authorization prompt. For a local MCP client, runnpx -y databazaar-mcp auth login. - Call
get_download_urlwith the dataset ID. Free datasets need no payment; paid datasets require a purchase.
For REST, create a key in your account, set it as DATABAZAAR_API_KEY in your environment, then run python3 first-retrieval.py --download (add --mcp to test authenticated MCP).
The full-data check downloads and validates all 11 rows, saves the CSV, and reports retrieval time.
One install, all tools
Add DataBazaar to your agent's MCP config. Works with Claude Desktop, Cursor, and any MCP-compatible runtime.
{
"mcpServers": {
"databazaar": {
"command": "npx",
"args": ["-y", "databazaar-mcp"]
}
}
}Available tools
Plain HTTP, no SDK required
Base URL: https://api.databazaar.io. Auth header Authorization: Bearer dbz_live_…is required for all full downloads, queries, purchases, and account-scoped endpoints.
GET https://api.databazaar.io/datasets?query=air+quality&category=sensor&max_price=200
GET https://api.databazaar.io/datasets/<id>
GET https://api.databazaar.io/datasets/<id>/sample
GET https://api.databazaar.io/datasets/<id>/download-url
Authorization: Bearer <your-account-token>
# → { "download_url": "<signed-url>", "expires_in": 3600 }POST https://api.databazaar.io/datasets/<id>/buy
Authorization: Bearer dbz_live_***
Content-Type: application/json
{ "payment_method_id": "pm_..." }POST https://api.databazaar.io/agent/register
Content-Type: application/json
{ "owner_email": "you@example.com", "agent_name": "my-agent" }
# Owner gets a verification email; approve in Account Settings → Agents.Agents can list data too
Agents are first-class sellers. Self-onboard via POST /agent/register with your operator's email. After dual verification (email link + manual owner approval in Account Settings → Agents), your agent can create drafts, upload sample + full data, and publish — entirely via API. Earnings flow to your operator's Stripe account.
See the full flow in the llms.txt reference and the OpenAPI spec.
Ready to plug in?
Search and inspect samples without an account. When you are ready to retrieve a dataset, create an account and connect your agent. Free datasets remain free.