Get freeFree trades

    Prices are Kalshi's current asks in cents. A 70¢ contract risks 70¢ to make 30¢. Heuristics, not odds of winning. 1-800-MY-RESET

    1. Home
    2. Markets
    3. Crypto
    4. BTC
    5. BTC 15-minute

    Kalshi · crypto up or down · 1m to 1W charts

    Bitcoin 15 Min Up or Down

    Kalshi opens a new up-or-down Bitcoin window on every quarter hour. This page is that contract: the price to beat, what Up and Down cost right now, and your payout after fees. The chart tabs change the candles, never the contract.

    BTC hourly deskHow to read itLesson 05
    BTC / USD Coinbase · 5-minute candles

    Short-term read

    Move needed
    —
    Typical move
    —
    ATR 14
    —
    15m range
    —
    1h range
    —
    vs VWAP
    —

    Heuristics, not odds of winning. σ = recent realized volatility of these candles, scaled to the time left.

    Coinbase BTC-USD candles, your time zone. Kalshi settles on the CF Benchmarks Bitcoin real-time index, which includes Coinbase. Charts by TradingView

    Last 24 windows

    Settled Kalshi 15-minute results. Most recent at the top.

    Kalshi’s settlement value. The chart above is Coinbase.

      Tap a row to mark that window on the chart. Final = Kalshi’s settlement value. Every settled window, day by day

      Settlement

      What price actually settles a 15-minute window

      Kalshi settles this 15-minute window on the CME CF Bitcoin Real Time Index, a CF Benchmarks real-time index. In the final minute before the window closes, sixty one-second prints of that index are collected and averaged, and that average is compared with the same sixty-second average taken before the previous close. The strike type is greater or equal, so a window that lands exactly level resolves YES.

      CME CF Bitcoin Real Time Index

      Kalshi

      Built from
      Order books, not trades. The constituent books are consolidated, uncrossed, and reduced to a depth-weighted mid.
      Updates
      About once per second, to roughly every 200 milliseconds on request.
      Administrator
      CF Benchmarks Ltd, an FCA-authorised benchmark administrator, register 847100.
      Methodology
      Published and versioned, with a named exchange list. CF calls this one a Real Time Index.

      The seven exchanges behind the number

      CoinbaseKrakenBitstampGeminiBullishCrypto.comLMAX Digital

      itBit was a constituent until 27 July 2026. Bitfinex and OKCoin were suspended back in April 2017 and have not fed the index since. Binance is not a constituent, which is worth knowing if you watch Binance and trade Kalshi.

      Sources: CME CF Constituent Exchanges v13.7 (24 Aug 2026), CME CF Real Time Indices Methodology, Kalshi on crypto markets. Read 26 Sep 2026.

      Chainlink TWAP 60s

      Polymarket

      Built from
      Three or more price aggregators and market data vendors using exchange order books. Chainlink does not name them publicly.
      Reported as
      A consensus median across the oracle network, delivered as a signed report.
      In use since
      7 August 2026 on Polymarket’s bitcoin up-or-down market, replacing a single-price snapshot.
      Methodology
      Not published. Polymarket’s own docs say the sampling boundaries, weighting and rounding are not specified.

      Polymarket warns readers not to try to reproduce the value independently “without a specification from Chainlink.” That is the honest difference between the two: one publishes its rules and its venue list, the other publishes neither.

      Ties work the same way at both. Polymarket resolves Up when the closing TWAP is greater than or equal to the opening price, exactly as Kalshi resolves YES on a level window.

      Sources: Chainlink Data Streams sources, Polymarket Chainlink TWAP docs. Read 22 Sep 2026.

      Do they disagree? Rarely, and almost only on near-flat windows. Both are order-book derived and both average about sixty seconds, so the outcome only flips when a window’s net move is smaller than the gap between the two indices, which research published on 24 August 2026 measured at roughly 2.5 basis points between a major exchange mid and the Chainlink price. That was measured on bitcoin, where it came to about $21 with bitcoin near $86,000. We have found no published study measuring the CME CF Bitcoin Real Time Index against Chainlink directly, so we are not going to put a number on that. The study.

      The record

      Same rule, same day, one side needed six times the bankroll

      On September 20, 2026 all 96 windows settled. Buy one side every window, $100 to start, double after every loss. Both sides finished ahead. What separates them is not the profit, it is what the run demanded of you before it got there.

      Always UP

      Bought YES every window

      Finished
      +$3,881.23
      Bankroll it needed
      $800.00
      Worst drawdown
      $700.00

      Biggest single stake $800. Longest losing run 3 windows. Its worst fall from a high came at 5:45 AM PT.

      Always DOWN

      Bought NO every window

      Finished
      +$3,207.41
      Bankroll it needed
      $4,982.59
      Worst drawdown
      $3,123.12

      Biggest single stake $3,200. Longest losing run 5 windows. At 9:45 AM PT it was $1,782.59 below where it started, before any of the profit existed.

      Five losses in a row is what a doubling rule actually costs. The DOWN run finished ahead, and to get there it had to put $3,200 on a single 15-minute window and fall $3,123 from its own high. If your account had held $3,000 that morning, the rule would have stopped working before the day turned. Read the bankroll column before the profit column, every time. Fees are inside every figure on that page.

      Days on the record

      Every published session

      The seven most recent of 45 published sessions, newest first. Across those seven sessions, 316 of 668 settled windows closed up and 352 closed down, 47.3% up. Every window is priced on its own, so a week of results is a record of what happened and not a signal about the next one. A session is one Pacific calendar day, midnight to midnight, and a complete day is 96 windows. A day showing fewer is either one the archive did not finish recording or one where Kalshi never listed those windows, and that day's own page says which. Neither is a short session.

      Questions

      What people ask about this market

      What price does Kalshi use to settle bitcoin markets?

      Kalshi settles BTC contracts on the CME CF Bitcoin Real-Time Index, administered by CF Benchmarks, an FCA-regulated benchmark administrator on register 847100. Per Kalshi’s help centre, crypto contracts settle by averaging sixty seconds of the Real-Time Index. Sixty one-second prints are collected in the final minute before expiration and averaged.

      Is Kalshi’s bitcoin price the same as Coinbase’s?

      No. Coinbase is one of seven constituent exchanges, alongside Kraken, Bitstamp, Gemini, Bullish, Crypto.com and LMAX Digital. The index consolidates their order books, uncrosses the result and takes a depth-weighted mid, so no single venue sets it. Binance is not a constituent, and itBit was suspended on 27 July 2026.

      What happens if bitcoin closes exactly at the price to beat?

      It resolves YES. KXBTC15M uses a greater-or-equal strike type, so a window that lands exactly level goes to the up side. Polymarket works the same way on its bitcoin up or down markets, resolving Up when the closing average is greater than or equal to the opening price. Ties favour up at both venues.

      Do Kalshi and Polymarket use the same bitcoin price?

      No. Kalshi uses CF Benchmarks’ Real-Time Index, a depth-weighted mid of seven named exchanges’ order books updated about once per second. Polymarket has used Chainlink’s BTC/USD 60-second TWAP data stream since 7 August 2026, sourced from three or more aggregators Chainlink does not name. The two typically sit within a few basis points.

      Is this the same index CME uses to settle bitcoin futures?

      No, they are different products from the same administrator. CME futures settle on the Bitcoin Reference Rate, a once-daily rate built from completed trades between 15:00 and 16:00 London, split into twelve five-minute partitions. Kalshi uses the Real-Time Index, the continuous order-book one. People conflate them constantly.

      Can two venues disagree about whether the same window closed up?

      Yes, though rarely. Both indices are order-book derived and both average about sixty seconds, so they usually agree. Research published on 24 August 2026 found a major exchange mid sitting about 2.5 basis points from the Chainlink price, roughly $21 with bitcoin near $86,000. Disagreement concentrates in near-flat windows.

      Reset monitor · this device

      Before the next trade

      After a loss, a big win or a long session, take a few minutes off the desk before the next trade.

      Why the break?

      On the desk today: . Counts while a desk page is open and on screen.

      Foreword

      Exclusive TunaLead writer · pen name

      Trading 15-minute markets moves fast. A setup can change in minutes, so I want the right information in front of me before I make a decision.

      That’s why I use PM Countdown. It puts the indicators and market information I care about in one place, so I can review the setup, understand the price being offered, and decide whether it’s actually worth taking.

      The reset is part of my process too. After a big win, a tough loss, or too much time at the screen, five minutes away helps me come back focused instead of chasing the next trade.

      I don’t need to trade every market. I want the right setup at the right price. PM Countdown helps me stay disciplined and wait for it.

      Exclusive Tuna

      Reset running

      5:00

      Answer honestly. Only you see this, and nothing is saved.

      1. Is the next trade about getting that loss back?
      2. Do I know today’s loss limit in dollars?
      3. Have I reached it?
      4. Would I size the next trade bigger than I planned this morning?
      1. Have I moved some of it to my bank account?
      2. Does money I just won feel easier to risk than money I put in?
      3. Would I size up the next trade because I’m up?
      4. Do I know when I’m stopping today?
      1. Have I stood up in the last hour?
      2. Have I eaten something in the last few hours?
      3. Did I get a normal night’s sleep?
      4. Is it later than I planned to stop?

      After a loss

      Let this one go.

      After a big win

      Move some of it to your bank.

      What to do with a big win: the set-aside, the move off the exchange, and the size of the next bankroll.

      Been at it too long

      Get up for a bit.

      Tools for a longer break
      Your limits

      The most you’re OK losing today. We’ll keep it in front of you.

      Saved only in this browser. Nothing is sent to us. Gambling problem? Call or text 1-800-MY-RESET.

      How to read this desk

      Four numbers, then a decision

      Every tab ends in the same question: is the price you'd pay below what you think the contract is worth, after the fee?

      Lesson 05: 15-minute markets
      1. To beat

        The strike. A 15-minute contract's strike is set when its window opens. An hourly contract has one market per strike, so the desk starts on the strike nearest the current BTC price and lets you pick another.

      2. Price now and the gap

        The live Coinbase price and how far it sits above or below the strike. Kalshi settles on CF Benchmarks' BRTI, which Coinbase feeds into, so the two track closely but aren't identical.

      3. Up and Down prices

        Up is the Kalshi ask for YES. Down is 100 minus the YES bid, which is what NO costs you. A price in cents reads as the market's chance in percent.

      4. If you're right, if you're wrong

        What your stake buys at the current ask, after Kalshi's taker fee. If you're wrong, the whole stake is gone. That's why the stake is the first thing to decide, not the last.

      15m

      Real contract. KXBTC15M opens a new window on every quarter hour and settles on the 60-second BRTI average at the close.

      1h · 4h · 1D · 1W

      Chart views. Longer candles on top of the same 15-minute contract. Kalshi’s hourly KXBTCD strike ladder lives on the BTC hourly desk, where you pick the hour and the strike.

      5m

      Our layout. The last 5 minutes of the live 15-minute window, shaded, with a countdown to when that phase starts.

      30m · 45m

      Our layout. Two or three 15-minute windows in a row. Each leg is a separate trade; a future leg's strike reads "Strike set when the window opens" until Kalshi sets it.

      1m

      Chart view. 1-minute candles on top of the current 15-minute contract.

      ETH · SOL · XRP · DOGE

      Real contracts. KXETH15M, KXSOL15M, KXXRP15M and KXDOGE15M work like KXBTC15M: a new up-or-down window every quarter hour, each settling on a CF Benchmarks index. The hourly strike ladder we cover is BTC only (KXBTCD); every chart tab works for every coin.

      Fees on fast markets

      A coin flip needs a real edge

      Most 15-minute windows trade near 50¢, and that's where Kalshi's fee is highest. The spread and the fee have to be beaten before you make anything.

      Worked example: 100 Up contracts at a 51¢ ask

      Cost is $51.00 plus a $1.75 fee (0.07 × 100 × 0.51 × 0.49, rounded up to the cent), so $52.75 in total. They pay $100 if BTC finishes above the strike. You need to be right more than 52.75% of the time just to break even, and selling early adds a second fee.

      Fee formula: Kalshi's taker fee, 0.07 × contracts × price × (1 − price), rounded up to the cent; resting limit orders can pay less. predictionmarketspicks.com calls the spread plus fees on a near-coin-flip contract "a steep, constant drag". TheLines puts the fee at 1¢ to 2¢ per contract.

      Sizing

      ¼ Kelly, capped at 5%

      Full Kelly for a contract at price P when you believe the chance is q is (q − P) ÷ (1 − P). We use a quarter of that, never more than 5% of your bankroll, and nothing at all when the number is zero or negative.

      Full Kelly—
      ¼ Kelly—
      Stake—

      Ignores fees, which lower it further. Your chance is a guess; if it's wrong, Kelly oversizes. Source: betstamp on Kelly, which advises never risking more than 5% on one bet.

      Take a break between windows

      A new window every 15 minutes isn't a reason to trade every one

      Fast markets reward slow rules. Decide how many windows you'll trade and your loss limit for the session before the first one, and stop when either runs out. If you lost the last window, sit out the next one. Our daily results archive plays a whole settled session back window by window, including what doubling after a loss demanded of the account.

      • Set a money limit for the session before you start, and keep it.
      • Never raise the stake to win back the last window.
      • Skip any window where you can't say why the price is wrong.
      • Walk away after a set number of windows, win or lose.
      Lesson 05: 15-minute marketsLesson 04: loss limitsSee a full day of settled windows

      TheLines' guide to these markets makes the same point: "Take breaks. Avoid rushing into consecutive trades."

      Only risk what you can afford to lose. This desk is information, not advice, and PM Countdown isn't affiliated with Kalshi. Kalshi users can set funding caps and opt-outs in the Responsible Trading Hub. Gambling problem? Call or text 1-800-MY-RESET. Responsible gambling guide