OddsRadarWire API Review 2026

OddsRadarWire is a sports odds API that delivers live and prematch prices from one upstream source: Sportradar's Betradar Unified Odds feed, as carried through a licensed operator. It is not affiliated with Sportradar or Betradar, and it says so on its own site.

Two things make the pitch clear on first contact. Every selection carries the provider's raw no-vig probability alongside the decimal price and a fractional representation, so you get the implied fair value without computing it yourself. And the demo key is genuinely public: key=demo works on the REST endpoints with no signup, which puts you from reading about the feed to seeing its JSON in under a minute.

The service comes from the same team that runs PinnWire (Pinnacle odds) and OddiWire (Oddin esports odds). All three share a design pattern: REST plus SSE, a public demo key, an llms.txt setup guide for AI coding assistants, and a permanent free tier.

The homepage leads with soccer, tennis, basketball and cricket, but the actual board is wider. We pulled the /v1/counts endpoint through the demo key and got live fixtures in tennis, table tennis, soccer, basketball, cricket, ice hockey, volleyball, and baseball. Upcoming fixtures added American football, futsal, and handball. The docs also reference CS2 (esports) and darts as sport filter values.

This is a Betradar board, so coverage follows whatever that upstream operator carries. During the early NFL and NHL seasons, expect American football and ice hockey fixtures to appear once regular-season games are on the schedule. The breadth you see depends on what the upstream book is trading at the time you look.

Markets are tiered by stability. Tier 1 covers the core lines (match winner, main handicap, main total, team totals, and their first-period equivalents, plus esports map markets) with stable canonical keys that won't change under you. Tier 2 adds player props and alternate lines in beta. Tier 3 is everything else, with raw provider market names and no stability guarantee. We verified Tier 1 canonical keys live through the demo: match_winner, main_handicap, main_total, and team_total all came back on soccer fixtures.

One thing to know before you design around it: this is a single-source feed. There's one price per selection, one upstream provider (Betradar), and no bookmaker field. You're getting one book's view of the market, not a multi-book comparison. For line shopping across named sportsbooks, pair it with an aggregator.

REST endpoints live under /v1/. The main one is /v1/fixtures, which returns the full board filtered by sport, event type (live or prematch), market status, tier, and canonical key. A single-fixture endpoint takes an event ID. /v1/counts gives fixture counts per sport. /v1/health reports upstream freshness and board size. /v1/me shows your plan, limits, and usage. Auth is a header (x-api-key or Authorization: Bearer) or a query parameter.

Each fixture carries competitors with abbreviation and country code, a start time, league and category strings, and a score block with per-phase breakdown. Each selection inside a market includes a decimal price, a fractional price, the provider's no-vig probability, a line and side for totals and handicaps, and a status (TRADING, SUSPENDED, or RESULTED). By default the API returns only OPEN markets with priced selections, so you don't have to filter out dead markets yourself. Pass include=closed,suspended,resulted if you want the full picture.

The SSE stream at /v1/stream pushes an odds event for every upstream price change, rebuilt from the current fixture state. It sends keepalive pings every 15 seconds. The stream uses the same normalized market set as REST, so parsing code works for both transports. SSE is available on the Full plan only ($299/mo). There is no WebSocket endpoint; the real-time push is SSE.

The service provides an llms.txt at the site root with a machine-readable API reference, and the homepage offers a setup prompt for AI coding assistants. There is no MCP server as of this review.

Three self-serve tiers, plus a comparison card for direct Sportradar access:

• Free ($0): 100 REST requests per day, the full board across all sports and tiers. No card required, no expiry. The public demo key (key=demo) is even lighter: 10 requests per 5 minutes, shared globally, no signup at all.
• Basic ($29/mo): 5,000 REST requests per day. No SSE, no drops.
• Full ($299/mo): 30 REST requests per second (no daily cap), unlimited SSE streaming, and unlimited drop alerts.

Card and crypto payments are both accepted. The gap between Basic and Full is large: $29 to $299 with nothing in between. If your product needs the live SSE stream, the entry price is $299, not $29. Basic is built for polling workflows and prototypes that outgrow the free tier.

The scope is deliberate, and the product makes most of it plain:

• Multi-book odds. One source, one price per selection. For price comparison across DraftKings, FanDuel, Pinnacle and the rest, run an aggregator from the odds API directory alongside it.
• Historical and closing odds. The API serves current prices only. For backtesting or closing-line analysis, pair it with a historical odds archive.
• Futures, deeplinks, bet settlement, same-game parlays. Markets are match-level: winners, handicaps, totals, and props (Tier 2 beta). No outright or season-long futures appeared in our testing.
• Scores beyond the score block. Each fixture carries a live score with per-phase breakdown, but there's no separate scores or stats service.
• An MCP server. The llms.txt file helps AI agents find the docs, but there's no packaged MCP server today.

This feed suits developers who want Betradar-quality odds in a self-serve package without a $30k/month direct contract:

• Odds dashboard builders who need a clean, normalized feed with stable market keys they can render straight to a UI.
• Price alert and drop-tracking tools that want SSE push events rather than polling, with the no-vig probability already attached.
• Prototype and hobby projects that can run on 100 free requests a day forever, testing against real Betradar markets.
• AI-assisted developers who want their coding agent to read the llms.txt, wire up the demo key, and return working code in one session.

Because it's a single-source feed, it pairs well with a multi-book aggregator for line shopping. Use OddsRadarWire for one sharp upstream price and an aggregator from the directory for breadth across books.

OddsRadarWire takes the Betradar Unified Odds feed and packages it as a clean, self-serve API at a fraction of the direct-provider price. The no-vig probability on every selection is a useful touch that saves a calculation step, and the tiered canonical keys give you a stable schema to build against without worrying about upstream name changes.

The public demo key means you can verify the data before handing over an email. The permanent free tier means a hobby project never pays. And if you need every price change as it happens, the Full plan's SSE stream delivers that without polling.

The trade-off is scope: one source, no history, no multi-book comparison. That's a clear shape, and the team is upfront about it. Build for that constraint, pair it where you need breadth, and the demo key tells you the rest in about two minutes.