PropLine API Review 2026

PropLine is a sports odds API built around player props. Most odds APIs treat props as a secondary market bolted onto game lines. PropLine inverts that: 60+ prop market types are first-class, with game lines (moneylines, spreads, totals) available alongside them. The API covers 54 sports across 15 sportsbooks and 5 exchanges, including Pinnacle, DraftKings, FanDuel, Bovada, and prediction markets like Kalshi and Polymarket.

The pitch goes beyond raw odds. PropLine resolves props against actual box scores and tells you whether a line hit, lost, or pushed. It calculates no-vig fair lines and flags cross-book +EV opportunities. It tracks line movement and runs steam detection. If your work starts with "which props have value right now," PropLine tries to answer that question inside the API itself rather than leaving the math to you.

PropLine lists 54 sports, though the practical depth varies. US majors (MLB, NFL, NBA, NHL, WNBA, NCAAF) get the deepest prop coverage. Soccer spans 28 leagues. UFC, boxing, tennis, and golf round out the list.

The 60+ prop market types are the differentiator. Baseball alone covers strikeouts, earned runs, hits, stolen bases, and batting stats. Basketball includes points, rebounds, assists, three-pointers, and double-doubles. Hockey has goals, shots, saves, and blocked shots. Period-specific markets (Q1, H1, P1, first-5-innings) are available through a query parameter rather than separate endpoints.

Book coverage is narrower than the big aggregators. You get 15 sportsbooks and 5 exchanges, not 300. The set tilts toward US-facing books and DFS platforms (PrizePicks, Underdog Fantasy, Sleeper). If you need a wide international book selection, this is not the right fit. If you need deep US props from the books that actually post them, the coverage matches the use case.

One useful detail: the response format is drop-in compatible with The Odds API's schema. If you already parse that format, switching costs are low.

Three features separate PropLine from a generic odds feed:

• Closing lines. The /closing endpoint returns opening and closing prices for resolved events, so you can calculate closing line value (CLV) without storing the feed yourself. Available from the Hobby tier ($9/mo) onward.
• +EV scoring. The /ev endpoint compares each book's price against a no-vig fair line anchored to Pinnacle or Bovada and flags positive expected value. Useful if you want the math done server-side.
• Steam detection. The /movement endpoint scores line moves on a 0 to 100 scale based on multi-book consensus. A high score means several books moved the same direction in a short window.

These are analytical features, not raw data. If you prefer to run your own models and only want clean odds, you can ignore them and use PropLine as a straight props feed. But they are the reason to pick this API over a broader aggregator that covers more books with less analysis. For more on how closing lines fit into a betting workflow, see the historical odds guide.

Pricing is published and starts at zero:

• Free: 1,000 requests/day. Game lines, scores, futures, stats, and context. No props history, no +EV, no closing lines.
• Hobby, $9/mo: 5,000 requests/day. Adds prop resolution, +EV scoring, line movement, closing lines, and 30-day history.
• Pro, $19/mo: 25,000 requests/day. Adds 90-day archive and CSV exports.
• Streaming Lite, $39/mo: 100,000 requests/day. Adds webhooks (HMAC-signed, with retry logic) and 180-day history.
• Streaming, $79/mo: 1,000,000 requests/day. Up to 10 webhooks and 365-day history.
• Enterprise: custom pricing for unlimited access.

Annual billing saves two months. A one-time $99 backfill option adds 7 days of historical tick data to any tier.

The free tier is genuinely usable for prototyping. It covers game lines, stats, and event context at 1,000 requests per day, which is enough to build a working integration and see the response shape. The analytical features (the reason you would pick PropLine specifically) start at $9/mo on Hobby, which is a low threshold to test whether the +EV and closing-line endpoints add value to your workflow.

The API lives at https://api.prop-line.com/v1. Authentication works through either a query parameter (?apiKey=) or a header (X-API-Key). Every response includes quota headers (X-Daily-Limit, X-Daily-Used, X-Daily-Remaining) so you can pace yourself without guessing.

Real-time delivery comes two ways. Webhooks: you register a callback URL with event filters (line movement, resolution, market suspension, steam) and PropLine pushes HMAC-signed payloads to it, batched up to 500 events per POST. Filters cover sport, event, market, player, and, since August 2026, bookmaker, so a consumer tracking three books receives only those books' moves. The retry policy escalates from 10 seconds to 1 hour across 6 attempts. And since late August 2026, a WebSocket stream at wss://ws.prop-line.com/v1/stream carries the same events with a per-subscription sequence number, so a client can detect gaps and replay. Both require the Streaming Lite tier or above.

On the ingest side, PropLine now consumes five books' own real-time feeds (DraftKings, FanDuel, Fanatics, Kalshi, and Polymarket) instead of polling them, and reports sub-second median push lag on those books. The rest of the board updates on a 60-second pre-game and 30-second live polling cadence. That mix is fine for props analysis; latency-sensitive arbitrage on game lines outside the streamed books is still a stretch.

Same-game parlay pricing arrived in September 2026. POST /v1/sports/{sport}/events/{id}/sgp takes 2 to 10 legs from one game and returns the book's own correlated price for that slip, alongside the independent product of the single-leg prices and the implied correlation factor. Legs use the same names as the odds endpoint, so building a slip is a copy-paste from an odds response. Three books price SGP at launch: FanDuel, BetOnline, and LowVig (the latter two share a bet-builder engine, so they quote the same number). Available on the Hobby tier and up, and wired into the SDKs, the MCP server, and the CLI.

SDKs exist for Python (pip install propline) and Node.js (npm install propline), plus a CLI tool. PropLine also ships an official MCP server with 26 tools covering live odds, +EV detection, graded results, scores, box-score stats, and line history. A hosted endpoint at https://mcp.prop-line.com/mcp works with no install and no key (a shared demo key provides free-tier data; pass your own key in an Authorization: Bearer header for full access). The npm package (npx -y propline-mcp) is still available for local use.

We ran a Streaming-tier key through several hundred requests in August 2026: MLB player props across a full slate, NFL week-1 props, futures for all six US sports, and the complete webhook lifecycle. The headline result is price fidelity. Every shared prop price we checked on Fanatics, DraftKings, and FanDuel matched independent reference feeds exactly, 100% accuracy across two sports. Coverage also runs ahead of the docs: the API served bookmakers the documentation does not yet list, and around 25 books quoted the MLB slate we pulled.

The webhook system held up under real traffic. Deliveries arrived as HMAC-signed batches every couple of seconds, thousands of movement events over a few minutes, with zero signature failures and per-element ids that make deduplication trivial. Integration took an afternoon, which is rare at this price point.

One finding turned into a correction that speaks well of the data model. Some BetMGM prop prices looked hours stale next to a reference feed, and we reported it as a capture bug. It was not one. BetMGM pulls player markets when MLB lineups drop, and PropLine keeps serving the last price with a suspended_at timestamp on the market, cleared when the book relists. We verified the timeline on our own examples down to the minute. The reference feed served those same withdrawn markets with no flag at all. PropLine was the only source in our test that could tell a live price from a pulled one.

The team also ships. Within a day of our feedback they added worst-case staleness and suspended-market counts to the freshness endpoint and a per-book filter on webhook subscriptions. We verified both live: the filter cut our delivery volume from roughly 15 posts a minute to 1.

Two practical notes remain. Player naming varies slightly between books, so plan a small reconciliation step. And we did not put the graded prop resolution through a full settlement cycle this session; that follow-up test is planned for early in the NFL season.

The product is built for a specific slice of the market:

• Props modelers who build projections and need cross-book prop prices with resolution data to grade their accuracy.
• +EV bettors who want server-side edge detection rather than building the no-vig math themselves.
• DFS players who need PrizePicks and Underdog pricing alongside sportsbook props, with payout multipliers and breakeven rates included.
• CLV trackers who measure their betting performance against closing lines and want that data from the API directly.

PropLine is happiest when props are the main course. If your product also needs broad international game-line coverage or ultra-low-latency arbitrage, it pairs naturally with a general aggregator from the sports odds API directory; plenty of teams run both.

PropLine is a props-first API that does more analysis than most odds feeds. Closing lines, +EV flags, steam detection, and graded results are built into the endpoints rather than left as an exercise for the developer. The pricing is low enough to test the analytical features before you commit.

The focus is the point. Coverage centers on the US books and exchanges that matter for props, the polling cadence is tuned for prop markets, and the historical archive grows deeper every day. If props matter to your product, PropLine is an easy recommendation: start on the free tier, pull a slate, and see the graded results for yourself.

What stood out when we vetted PropLine for the directory:

• 100% price accuracy against independent feeds in our testing
• Signed, batched webhook push with per-book filtering, integrated in an afternoon
• Shipped two of our feature requests the day we reported them

Tested on a live Streaming key in August 2026: prop prices matched independent feeds with 100% accuracy on the books we checked, and the webhook stream ran fast and flawless. When we flagged gaps, the team shipped fixes the same day, including per-book webhook filters and per-market staleness fields. A props-first feed that over-delivers for the price.

Last reviewed: 2026-08 · How providers get listed