Getting started
Before you start, create OAuth 2.0 credentials (a client_id / client_secret) for your white label in the Betting Backoffice. You exchange them for a short-lived access token that authenticates your API requests.
Note
Read more info within the Authorization section.
Common integration with the DATA.BET Sportsbook can be described using the next diagram:
Connecting to API
Every request to our API must carry a Bearer access token obtained from your OAuth 2.0 credentials.
For detailed information on how to obtain a token and authenticate requests, including curl examples, refer to the Authorization documentation.
IP whitelist
On top of authentication, the Betting API is restricted to a per-whitelabel allowlist of client IP addresses. A request that authenticates correctly but originates from an IP that isn't on your list is rejected with HTTP 403 Forbidden. Before you start making requests, make sure your outbound IPs are whitelisted.
You manage this list yourself through the Betting Backoffice — see the IP Whitelist guide.
MCP server
The Betting API is available as an MCP ⤴ (Model Context Protocol) server. Connect an MCP-capable AI tool to it and explore the API conversationally instead of assembling requests by hand. It can search the published API documentation directly — list the available specs, look up an operation's request schema, or list the callbacks you need to implement — and it can call the live API on your behalf, for example to look up a dictionary value, create a freebet, or check a bet report.
Integration environment only
This server calls the Betting API on betting.int.databet.cloud, the integration environment (see Authorization). It is not intended for production traffic. Use it to explore the API and validate your integration, then move to the production Betting API for real bets.
What it exposes
The server gives an AI assistant two kinds of tools.
Documentation tools — query the published contract directly, no live API call involved:
get_contract— with no arguments, every bundled spec and who implements it; with a spec name, its operations; with a spec and an operation, that operation's full request schema, resolved.list_callback_endpoints— every callback endpoint you must implement, with the request it sends and the response it expects.build_callback_fixture— a spec-conformant example payload for a callback endpoint, built from the spec's own documented examples.
API tools — one tool per Betting API operation, bundled from every spec published in this docs section:
- Dictionaries — sports, tournaments, markets, and other reference data. See Dictionaries.
- Bonus mechanics — Free Bet, Bet Insurance and Bet Boost.
- Bet actions — see Bet Actions.
- Bet reports — see Bet Reports.
- Player risk scores — see Player risk scores.
- Session tokens — the same tokens used by the Betting Frame; see Token.
- SPA integration metadata — server-side rendering initialization and static asset endpoints the Betting Frame uses.
Banners and the bet, cash out, insurance and heartbeat callbacks are not called this way — you implement them, and the platform calls you. Discover their contract through list_callback_endpoints and build_callback_fixture above.
If an operation is not documented in this Betting API section, the server does not expose it either — it has no tools beyond these specs.
Example
Once connected, ask your AI tool things like:
- "What specs does this server know about, and who implements each one?" — calls
get_contractwith no arguments. - "Show me the request schema for
POST /bet/place." — callsget_contractwith the spec and operation. - "What callbacks do I need to implement, and what do they expect back?" — calls
list_callback_endpoints. - "Give me an example payload for the cash out callback." — calls
build_callback_fixture. - "What sports and markets does the Betting API support?" — pulls from Dictionaries.
- "Create a €10 freebet for player 12345, valid for 7 days." — calls the Free Bet API.
- "Show me yesterday's bet report." — calls Bet Reports.
- "Cancel bet #98765." — calls Bet Actions.
- "What's player 12345's risk score?" — calls Player risk scores.
- "Generate a demo session token in EUR." — calls Token.
The assistant picks the right tool, fills in the request from the conversation, and shows you the response — you can inspect it, adjust the request, or ask a follow-up.
Prerequisites
You need a client_id / client_secret pair for your white label. Create one yourself in the Betting Backoffice — see OAuth2 Credentials.
Connecting
The server is published at:
1 | |
Open claude_desktop_config.json and add an entry to mcpServers:
1 2 3 4 5 6 7 8 9 10 11 12 | |
Restart Claude Desktop to pick up the change.
1 | |
Any client that can spawn a local MCP server process can connect the same way — configure it to run:
1 | |
A client with native support for remote (HTTP) MCP servers can instead connect directly to the URL above, without mcp-remote.
Authorizing
On first connection, mcp-remote opens a browser tab to a page titled Authorize API access, with two fields — Client ID and Client Secret. Enter the credentials from Prerequisites and click Authorize. The tool then stores a refresh token, so you will not be asked to authorize again until the token expires or is revoked.
If revoking a client_id / client_secret in the Betting Backoffice, disconnect and reconnect the tool afterward — see Revoking a credential.
Backend integration
Token and session management
The first step on your backend side is the ability to generate betting tokens for your users' sessions. You can Heartbeat Callback API to manage your sessions.
For more details about token generation, session lifecycle management, and typical session flow, refer to the Token documentation.
Bet Callback API
The Bet API is the core integration component that must be implemented on the client side. This component handles the critical betting callbacks and interactions.
For more details, refer to the Bet documentation.
Optional integrations
The following features are optional but commonly implemented by partners:
Free Bets and Bet Insurance are bonus mechanics. Partners usually integrate these into their gamification or bonus engine (e.g., welcome bonuses, achievement rewards, tournament prizes, etc.).
To create bonuses, you need access to dictionary entities (sports, leagues, markets, etc.) to specify your bonus conditions. For more details, refer to the Dictionary API.
Admin use cases: Bet Actions API
Partners may want to:
- Fetch betting history for their admin panels.
- Cancel or void specific bets manually.
For that purpose, the following Bet Actions is provided
Callback server requirements
Once you're ready to start receiving heartbeat/bet callbacks, provide your callback server URL.
Tip
You can find a simple example of how to develop your own callback server using Node.js in the GitHub repository databet-cloud/callback-server-example. It utilizes every type and callback described within the Bet Swagger specification page.
Frontend integration
Once you are able to create a valid betting token, you can embed the SPA directly into your website with the default view.
Quick start
For testing, you can even create a token using a simple curl command.
For production, your backend must generate and manage per-user tokens.
Adding the bootstrap script
To initialize the SPA, include the bootstrap script on your page.
For more details, refer to the Integration Bootstrap Script section.
This script loads the application and handles communication with the backend.
Add required DOM containers
You'll need to add specific DOM containers where different SPA components will be rendered.
For instructions, refer to the SPA Integration section.
After adding the containers and calling the bootstrap logic with the generated token, the SPA will be fully embedded and functional.
Customization of layout, routes, theme and widgets
For customization of layout, routes, theme and widgets, the Data.Bet team will be able to guide you through the process.
Integration testing before go-live
Before go-live, a mandatory integration testing phase will be run.