Home/Guides

How to choose an odds feed

How do I compare odds feeds before paying for one?

Choose an odds feed by what it does on your own markets, not by the speed it advertises. Check that it carries the books licensed in your state on the markets you bet, see how it bills, then record a trial window from each candidate and compare the recordings. Feeds are sold on a published latency figure, and several publish none at all; neither tells you what the feed does at the minutes you actually bet.

Free. No account and no subscription. Needs macOS 14 or later on a Mac with Apple silicon (M1 or later). 34.9 MB, notarised by Apple.

The honest way to choose is to record a trial window from every candidate and replay those recordings through the same engine. Same markets, same clock, same decision rules. Only the feed changes. A quoted latency is a claim about the vendor. A replay is a measurement of your book.

What to compare first

Start with what you can bet. A feed is worth paying for only if it carries the books licensed where you live, on the markets you bet: main lines, alternate lines, player props. Ask for the list by state, not the headline count of bookmakers.

Then how it charges. Some feeds bill per request, some per object returned, some in credits that depend on what you ask for. The same refresh can cost very different amounts on each.

Then how fresh it is, and whether it says so: the refresh interval, and a timestamp on each price saying when the feed saw it. The rest of this guide is about why the second matters more than the first.

Last, history. A feed that sells historical odds lets you replay a trial window before you commit to it.

What two public feeds charge

Both of these publish their prices. Many odds feeds sell only through a sales call.

FeedPlansHow it chargesRead
The Odds APIFree: 500 credits a month. Paid: $30 for 20,000, $59 for 100,000, $119 for 5 million, $249 for 15 million.Live odds cost 1 credit per market per region; historical odds 10. Listing sports and events is free.read 2026-10-16, source
SportsGameOddsFree: 2.5k objects a month, 10-minute updates, 9 bookmakers. $99: 100k objects, 3-minute updates, 77 bookmakers. $299: no object cap, sub-minute updates, 82 bookmakers.Objects returned per month, with a per-minute request limit.read 2026-10-16, source

A month of credits, worked

On a credit feed, one refresh costs the number of markets times the number of regions. Moneylines, spreads and totals for one NFL slate in the US region is 3 markets in one region: 3 credits. Add the EU region, which carries the sharp book a fair price is taken from, and it is 6.

The free plan's 500 credits a month cover 166 refreshes of the first or 83 of the second. Refreshing every 15 minutes through a 4-hour Sunday window is 16 refreshes and 96 credits, so the free plan covers 5 Sundays like that and nothing else. One historical snapshot of the same request costs 60.

Do that sum before you pick a plan. Bookbreaker does it beside its refresh button, before anything is fetched.

Replay, not datasheets

Replay only works if the recordings line up in time. Ours refuses to compare captures whose clocks disagree by more than 120 seconds, because past that you are measuring clock drift and calling it latency. It also applies a stated prior: a latency floor of 2.0 seconds, below which no feed is credited with being faster. That floor is a prior, not a measurement.

Why latency decides more than it looks

A quote arrives stamped at 10.0 seconds old. The feed carrying it took 5.4 seconds, above the stated floor. So the price you are pricing against is really 15.4 seconds old. Everything downstream moves with that.

Fill probability splits the same way. Read at stamped age, the fill looks like 89%. Read at effective age, it is 84%. On a nominal edge of 4.0%, the naive read books 3.58% and the honest read books 3.37%. The smaller figure is the one that settles. Both fills come off a stated prior survival curve for this market class, not off a measurement of any feed. Your own accept and reject record replaces it.

The table below shows the same mechanism on a different stated prior: survival modelled as exponential decay with a tau of 60.0 seconds, tabulated from the latency floor prior of 2.0 seconds upward. The half-life that follows from that tau is 41.6 seconds. Modelled, not measured, and it falls at every step.

Quote age (s)Edge surviving
296.7%
592.0%
1084.6%
3060.7%
6036.8%

Read any such curve against effective age, never against stamped age. An effective age of 15.4 seconds still sits inside the half-life; a stamped age of 10.0 seconds flatters the feed by exactly the latency you failed to subtract. Do not expect this table to reproduce the fill figures above: different tau, different curve. What carries across is the direction.

The thing nobody sells on

A feed that stamps its own observation time is worth more than a faster feed that does not. Without a stamp you cannot recover effective age, so you cannot compute the honest fill, so the 3.58% figure is the only one available to you and it overstates. Speed you cannot check is a marketing claim. A timestamp is evidence. Buy the evidence.

How old is the price on your screen →