Can an AI agent beat Kalshi's 15-minute Dogecoin market?

From a rule to a record

A backtest of a Dogecoin rule tells you what it would have done against prices already known. This page wants the other thing: how the idea that a thin early-morning book overshoots fares against the KXDOGE15M book as it stands, at :10, :25, :40 and :55 past each hour from 5 AM to 8 AM Eastern, filled at the ask, charged Kalshi's fee (about 1.4¢ a contract at 27¢), and settled when CF Benchmarks' DOGEUSDRTI says so. So the Dogecoin agent runs on Windmill like any user's agent, and each bet it places on KXDOGE15M is written here as it settles, wins and losses alike.

How the agent works

The agent is the rule below in plain English, on Claude Haiku 5.5, scheduled four times an hour from 5 AM to 8 AM Eastern. It builds its record in four steps.

1. Read the open window

At :10, :25, :40 and :55 past each hour from 5 AM to 8 AM Eastern, it reads the open KXDOGE15M window, with the live bid and ask on each side and the minute candles since it opened. Kalshi lists a new KXDOGE15M window every fifteen minutes, and settled 94.9 a day over the last seven complete days. It can also read Coinbase's DOGE-USD price as a stand-in for CF Benchmarks' DOGEUSDRTI, the index the window settles on; between runs nothing watches the market.

2. Filter for the setup

It places nothing unless this holds:

• one side's price rose 8¢ or more over the last 3 minutes (YES bid now against the 1-minute candle 3 minutes earlier) to an ask of 62–85¢

Why this setup: The by-hour table below has Dogecoin's thinnest windows in the 5 AM, 6 AM and 7 AM ET hours; the rule fades a late jump only then, on the idea that a thin book overshoots, and buys the cheap side, so most of its bets lose and a win pays for several.

3. Size and place

It buys 5 contracts of the other side at its ask, on paper against the live KXDOGE15M book: a paper order fills only against what the book shows at that price. At about 27¢ a contract, Kalshi's taker fee at KXDOGE15M's settings (fee type "quadratic", multiplier 1) is about 1.4¢. So the five contracts cost about $1.42 all in, and pay $5.00 if they win.

4. Hold to settlement and record it

Every Dogecoin position is held until its KXDOGE15M window closes, at most fifteen minutes after the order, and Kalshi settles it on CF Benchmarks' DOGEUSDRTI; nothing is carried past 8 AM Eastern. The price paid, the fee, the result and the net go into the record below as each bet settles. Most of the 12 runs a day end with no order, and the agent's log keeps those too.

The record: October 30, 2026

The agent started October 28, 2026. Two bets have settled since, all on October 29, 2026: won 2, lost 0.

The two settled bets, newest first:

Two settled bets is a start, not a result; nothing is read into the record before thirty.

It runs this page's starter rule, word for word:

“At :10, :25, :40 and :55 past each hour from 5 AM to 8 AM Eastern, on the open KXDOGE15M window: if one side's price rose 8¢ or more over the last 3 minutes (YES bid now against the 1-minute candle 3 minutes earlier) to an ask of 62–85¢, buy 5 contracts of the other side at its ask. Otherwise no order. Hold to settlement.”

Run this rule as a Windmill agent

The market the rule is betting into

Over the last 1,000 settled windows, from Oct 19, 2026 to Oct 29, 2026, Dogecoin closed at or above its start 482 times and below it 518. The longest run was ten NO windows in a row. The median move from start to close was 0.14% of price. Volume ran at a median of 5.2 million contracts a day; the thinnest hours, by a median window's contracts, close at 5 AM, 6 AM and 7 AM Eastern.

The number that matters most for any rule that buys the favorite is whether the favorite pays for itself. For each of the last 200 windows up to Oct 29, 2026 we read Kalshi's quote 5 minutes before the close and took the favorite, the side priced above 50¢, at its ask; 200 had a quote on both sides, and the favorite won 173 of them. Breakeven is the price plus Kalshi's taker fee, as a share of the $1 a winning contract pays.

Three notes on that table. The 80–89¢ band has the widest gap between win rate and breakeven, 88% against 85.6%, and at 43 windows its interval runs from 76% to 95%, which includes breakeven. Two of the five bands fell short of breakeven, the 60–69¢ band by the most: 65% against 67.7%. And every band's interval includes breakeven, which is why the last column reads the same five times. In our read of Kalshi's prices, the Dogecoin favorite 5 minutes before the close is priced about right. Nothing on this page says a buyer of the favorite has an edge.

The rule under test pays 15–38¢ a contract, below every favorite in the table: it buys the side the market is against, so the table prices the other side of its bets, not its own.

Kalshi's rules, in Kalshi's words

This is the rule of the market that was open (KXDOGE15M-26OCT292345-45) when we read Kalshi's API on October 30, 2026, quoted from Kalshi; the series and its full rules are on Kalshi's site. The times and the target change with every market:

“If the simple average of the sixty seconds of CF Benchmarks' DOGEUSDRTI before 11:45 PM EDT on Oct 29, 2026 is at least the simple average of the sixty seconds of CF Benchmarks' DOGEUSDRTI before 11:30 PM EDT on October 29, 2026, then the market resolves to Yes.”

The series names CF Benchmarks as its settlement source. Kalshi lists the series with the fee type "quadratic" at a multiplier of 1: a taker pays 1.8¢ a contract at 50¢ before rounding, less toward either end, and a resting order that fills later pays nothing. The fee calculator works out the fee and the breakeven for any price.

Dogecoin's windows settle on CF Benchmarks' DOGEUSDRTI averaged over the last sixty seconds at each end, and the value is rounded to 7 decimal places rather than the 4 of the XRP and Solana windows, since one dogecoin trades for well under a dollar. (Kalshi's API, the series' settled markets, checked October 28, 2026.)

What success looks like

A positive net per contract after Kalshi's fees over thirty settled bets, with the confidence interval clear of zero, would be a result.

A rule that only matches its price is not one. A contract bought near 27¢ that wins as often as that price says, 27 times in a hundred, loses about 1.4¢ a contract to the fee alone, so the line to beat on KXDOGE15M is a win rate of about 28.4%, not 27%. Each win at that price nets about 71.6¢ a contract after the fee and each loss costs 28.4¢, so one win covers 2.5 losses.

At most 12 KXDOGE15M runs a day means thirty settled bets take at least three days, and longer when the setup is rare. If the record climbs, the idea that a thin early-morning book overshoots holds at these prices. If it falls, the idea goes, and the record stays on this page.

Follow along

• This page refreshes every hour from Kalshi's settled data and the agent's own log.
• Copy the Dogecoin agent: the button below opens the builder with this KXDOGE15M rule, to run as written on paper under your own account, or with your own numbers in place of 15–38¢.
• The fee calculator shows the fee and the breakeven at any price on KXDOGE15M.
• The 90-strategy test is the longer version of this exercise: about 90 rules and 4,200 paper bets across ten days.
• The Kalshi markets hub lists the other 15-minute markets and every other market page.

Run this rule as a Windmill agent

Questions

What does Kalshi's 15-minute Dogecoin market settle on?

On CF Benchmarks: the rules quoted on this page, read on October 30, 2026, compare the price at the end of the window with the price at its start, and a tie resolves Yes.

What is the fee at 50¢?

A taker pays 1.8¢ a contract before rounding at 50¢ on this series (fee type "quadratic", multiplier 1), less toward either end. The fee calculator gives the fee and breakeven at any price.

Is a bot here profitable?

Nothing on this page says so. The favorite 5 minutes before the close won 173 of 200 windows in our read of Kalshi's prices, and the table shows each price band against its breakeven: 5 of its 5 bands are inside noise. Run your own rule on paper to see what it does, before you risk money.

Paper or live?

Agents trade on paper unless you build them live. Paper orders fill against Kalshi's live order book with Kalshi's fees and settle when the real market settles. To trade real money, connect your own Kalshi API key and build a live agent; your money stays on your Kalshi account.