Sports Odds API for AI Agents & MCP

Add live Australian odds to AI agents with the SportsGameOdds MCP server pointed at SockOdds, the five SGO SDKs, and LLM-ready docs.

Because SockOdds speaks the SportsGameOdds schema, the SGO tooling for AI works here: the sports-odds-api-mcp MCP server pointed at our base URL, the five SDKs, and LLM-ready docs (llms.txt, llms-full.txt, openapi.json, and a Markdown version of every page).

Why build your AI agent on SockOdds

An MCP server that already exists. Set its base URL to SockOdds; see AI-assisted development.
Grounding data. fairOdds and bookOdds per market.
LLM context. llms.txt, the full docs payload, the OpenAPI spec.
Language coverage. Five SDKs, one schema.
Markdown for every page. Append .md or send Accept: text/markdown.

How an AI agent works with our data

Install the MCP server (execute + search_docs tools; needs Deno)

claude mcp add sockodds --env SPORTS_ODDS_API_KEY_HEADER=… --env SPORTS_GAME_ODDS_BASE_URL=https://api.sockodds.com/v2 -- npx -y sports-odds-api-mcp

Ground answers in fairOdds and bookOdds

The agent reads a market and quotes a real price with a real timestamp.

Call from an SDK

Or from any HTTP client with the key.

Example request

curl "https://api.sockodds.com/v2/events/?leagueID=AFL&oddID=points-home-game-ml-home&includeOpposingOdds=true&limit=1" \
  -H "x-api-key: YOUR_API_KEY"

What an agent sees when a user asks who is favourite in an AFL game:

{
  "eventID": "afl_2026-09-03_fremantle_vs_hawthorn",
  "leagueID": "AFL",
  "odds": {
    "points-all-game-eo-odd": {
      "oddID": "points-all-game-eo-odd",
      "opposingOddID": "points-all-game-eo-even",
      "marketName": "Odd/Even Total Points",
      "statID": "points",
      "statEntityID": "all",
      "periodID": "game",
      "betTypeID": "eo",
      "sideID": "odd",
      "fairOdds": null,
      "bookOdds": "-116",
      "byBookmaker": {
        "unibet": {
          "odds": "-115",
          "decimal": 1.87,
          "available": true
        },
        "tabtouch": {
          "odds": "-118",
          "decimal": 1.85,
          "available": true
        }
      }
    }
  }
}

SockOdds vs wiring odds into an agent yourself

Capability	SockOdds	wiring odds into an agent yourself
Agent tool surface	`sports-odds-api-mcp` with the base URL set	Hand-write tool definitions
Grounding data	`fairOdds` / `bookOdds`	De-vig yourself
LLM context	llms.txt, llms-full.txt, OpenAPI	Write your own
Language coverage	Five SDKs	A client per language
Schema consistency	One `oddID`	Reconcile per source

Frequently asked questions

What MCP server does SockOdds use?

The SportsGameOdds one, with its base URL pointed at SockOdds.

How do I add Australian odds to an AI agent?

Install the MCP server or paste the AI context from the docs.

Which AI coding tools are supported?

Claude Code, Cursor, VS Code and anything that speaks MCP.

Can an LLM get accurate odds without hallucinating?

Yes, when it reads them from the feed with a timestamp.

Does the MCP server support streaming?

SockOdds does not stream; the tool polls.

Is there a free tier for prototyping?

Yes.

Related use cases

Odds API for Betting Alert Bots: Build a sports odds alert bot for Discord or Telegram: diff odds across every AU book to detect line moves and push deeplinks to users..
Sports Odds API for Betting Apps: The sports betting app API with odds from every major Australian bookmaker, event status, results, and click-to-bet deeplinks in a single feed..
Odds Comparison & Line Shopping API: Compare bookmaker odds with one API call.
Pricing: plans priced on request rate and scope, including a free Lite tier with no card.

Start building today

Free plan available. Set up in 5 minutes. Scale when you're ready.

Get your API key Explore the docs