Limbo casino game guide: how Stake Limbo works, what it pays and how to verify it

CodePACTBONUS
Claim the deal

200% bonus + 3.5% rakeback · 18+ · Wagering terms apply

Not the boy in the forest. The Limbo casino game is a Stake Original with one input: you name a multiplier, the game reveals one, and you are paid your target if the result is at or above it. We opened Stake Limbo on 11 October 2026, pushed the target field to both of its limits, read the auto-bet panel, the hotkeys and the Fairness tabs, and ran the published formula through a thousand bets at seven targets. Rules, maths, Limbo against Crash and Dice, verification, strategies that change nothing, predictors that cannot work, and how the Pactbonus deal applies.

What the Limbo casino game is

Not the Playdead game

Search for a limbo game and the results are a 2010 puzzle-platformer. This page is about the other Limbo: you type a target multiplier, the game reveals a result multiplier, and if the result is at or above your target you are paid stake times target. Below it, the stake is gone. That is the whole game, which is why a round settles in under a second.

We opened stake.com/casino/games/limbo on 11 October 2026. The panel shows an RTP of 99.00% with the tags Provably Fair and Stake Originals. Limbo is playable only on stake.com; the other Originals are in our Stake games review.

One round, step by step

  1. Set the bet amount; the ½ and 2× buttons halve or double it.
  2. Type a Target Multiplier or a Win Chance; the fields are linked and the game loads at 2.00× and 49.50%.
  3. Press Bet; the large 1.00× on screen flips to the round’s result.
  4. At or above the target, the bet pays stake × target (the Profit on Win box shows the figure beforehand); below it, the stake is lost.

The limits we measured

The target field accepts 1.01× to 1,000,000×: type 1.00 and it answers “Minimum is 1.01”, type 5,000,000 and it answers “Maximum is 1000000”. The linked win-chance field runs from 98.01980198% at 1.01× to 0.000099% at a million, 99 divided by the target every time. The description’s lines about results from 0.99× to infinity and payouts beyond 1,000,000× are copy, not rules: the published code floors every result at 1.00× and the field refuses a higher target.

The maths: Limbo RTP, win chance and variance

Win chance = 0.99 / target

The Limbo multiplier you choose is also the payout, and the win chance is 0.99 divided by it: 49.5% at 2×, 9.9% at 10×, 0.099% at 1,000×. Multiply the two and you get 0.99 at every target, so the expected return is 99 cents per dollar whether you ask for 1.01× or a million. A fair game would divide 1 by the target; the missing hundredth is the house edge. Limbo RTP is 99.00% because the formula says so, whatever the target.

Worked table: 1,000 bets of $1

Computed by us from the formula; the losing-streak column is the mean of 2,000 simulated sessions.

TargetWin chanceExpected result per $1Expected result, 1,000 betsStandard deviation, 1,000 betsLongest losing streak, typical (bets)
1.1×90%−$0.01−$10about $103
1.5×66%−$0.01−$10about $226
2×49.5%−$0.01−$10about $329–10
5×19.8%−$0.01−$10about $6326
10×9.9%−$0.01−$10about $9449
100×0.99%−$0.01−$10about $313285
1,000×0.099%−$0.01−$10about $995790

Every row costs the same $10 per $1,000 wagered; the rows differ only in spread. At 1.1× a thousand bets finish within a few dollars of −$10. At 1,000× the standard deviation is a hundred times the expected loss and 37% of thousand-bet sessions contain no win at all: $1,000 for a 63% chance at one $1,000 payout, same 1% edge.

Claim the deal with code PACTBONUS

Limbo vs Crash, and Limbo vs Dice

Same curve as Crash, without the timing

Limbo vs Crash is the comparison people make first, because the payout curve is identical: the chance that a Crash round reaches a multiplier m is 0.99 / m, Limbo’s win chance at target m. Everything around the curve differs. Crash is multiplayer: a shared round every few seconds, a rising multiplier, a live list of other people’s cash-outs and a button to press before the crash. Limbo is single-player and instant: no round to wait for, no climbing number, no cash-out timing, no bet list. Same 1%, less theatre, and several times as many bets a minute.

Dice in a different coat

Dice at 2.00× is Limbo at 2.00×: 49.5% win chance, the same payout, the same expected loss, the same spread. Dice presents the chance as a slider over a 0 to 100 roll and stops at 9,900×; Limbo presents the multiplier and runs to 1,000,000×. Stake’s own description calls Limbo a version of dice with no upper limit, which is fair. Plinko and Mines add a paytable and a choice, Keno a draw; Limbo adds nothing, and that is its appeal.

Stake Limbo vs other crypto Limbo games

Crypto Limbo outside Stake is rare. A handful of crypto casinos run a game under the same name, reported at 98% to 99% RTP, most without a published verification rule. The limbo game casino players search for is this one; if you find another, read its rule and its licence first.

Provably fair for Limbo, step by step

The published rule

Stake’s game-events page gives Limbo one float per bet and a three-line rule: divide the house-edge factor 0.99 by the float, floor the result to two decimals, and if it is below 1.00 make it 1.00. The float is HMAC-SHA256 of the server seed, your client seed and the nonce, turned into a number between 0 and 1 (our provably fair guide walks through it). The edge is visible in the code: the 0.99 is why 1% of floats land under 1.00× and pay nobody.

A worked example

For our Dice example we generated a server seed, used the client seed pactbonus and nonce 1, and Stake’s calculator produced the float 0.380899459356. In Limbo the same float gives 0.99 / 0.380899459356 = 2.5991…, floored to 2.59×: a 2.00× target wins and pays double, a 2.59× target wins by the smallest possible margin, a 2.60× target loses. The seeds are in the guide; enter them at stake.com/provably-fair/calculation with Limbo selected.

Checking your own bet

  1. Fairness, Seeds tab: type your own client seed before the first bet and copy the hashed server seed.
  2. Play. Each bet raises the nonce by one; the bet details show it.
  3. Rotate the seed pair; Stake reveals the old server seed. SHA-256 it: the output must equal the hash from step 1.
  4. Verify tab: choose Limbo, paste client seed, server seed and nonce. The result must match the bet.

That proves the result was fixed before you bet and not changed after; it does not prove the game is generous (the 0.99 settles that) or say anything about the licence, covered in is Stake safe.

Limbo strategy, honestly

Nothing moves the 1%

A Limbo strategy chooses a target and a staking pattern. The target sets the variance, as the table shows; the pattern decides when you lose, not whether. “Grind 1.01×” wins 98% of bets and loses the same $10 per $1,000 at a crawl; “lottery 1,000×” loses 999 bets in a thousand and the same $10 on average. Both are 99% in different clothes, and results are independent: a run of low results makes a high one no likelier.

Martingale at 2×

Doubling after each loss at 2× feels safe because each bet is close to a coin flip. Seven straight losses have probability 0.5057, about 0.84%, roughly once in 119 attempts; by then 127 units are gone and the next bet is 128, a $128 bet to win back $1 of profit. The bet limit, your balance or the welcome bonus’s max-bet rule (about 1% of the deposit) ends the chain first, and because the edge is charged on every doubled bet, the expected loss grows with the turnover. Paroli and D’Alembert reorder the same losses.

Limbo auto bet, hotkeys and Instant Bet

The Auto tab

The Limbo auto bet panel asks for a bet amount and a Number of Bets, where 0 means until you stop. The Advanced switch adds four rules: On Win reset the bet or increase it by a percentage, On Loss the same, Stop on Profit and Stop on Loss. Fill in Stop on Loss before you press Start Autobet, and read the increase-on-loss field as Martingale with a tidier name.

Hotkeys and settings

The gear menu under the game holds volume, Instant Bet, Animations, Max Bet, Game Info and Hotkeys. The hotkeys panel lists four keys for Limbo: space places the bet, s doubles it, a halves it, d zeroes it; unlike Dice there are no keys for the target. Instant Bet skips the result animation, and with both on a bet a second takes no effort.

Limbo predictor apps, bots and fakes

Why a predictor cannot exist

A Limbo predictor would need the next result, which means the server seed. It stays secret until you rotate it, and rotating starts a new pair, so the seed you can see never produces a result you have not already bet on. Apps and Telegram channels selling Stake Limbo “signals” or “seed hacks” are random-number toys at best and credential phishing at worst. Stake Limbo has no download; it runs inside stake.com.

Bots and scripts

A bot presses the button for you; every bet it places is still 0.99 / target against a result it cannot see, so its expected result is −1% of whatever it wagers, only faster. Scripts sold as profitable are the Auto tab with extra steps.

How the Pactbonus deal applies to Limbo

Welcome bonus and Originals weighting

Register with code PACTBONUS and the first deposit, $50 or more, is matched 200% up to $1,000: $100 becomes $300. Wagering is 40× deposit plus bonus, typically within 30 days, and Stake Originals count 100% toward it at the time of writing, so Limbo qualifies. The max bet with the bonus active is about 1% of the deposit: on $100 that is $1 bets, 12,000 of them to clear $300, at an expected cost of about $120. Conditions: bonus terms; the offer: the Stake deal.

Rakeback and VIP progress

Rakeback with the code is 3.5% of the house edge on every bet, for life, with no wagering. On Limbo the edge is 1%, so $100 wagered carries $1 of edge and returns $0.035; $10,000 returns $3.50. It never turns the 1% positive. Every bet also counts toward VIP progress and the $30,000 monthly Pactbonus leaderboard: VIP and rakeback.

Limbo is as fast as Dice: set the limits first

One field, one keypress, no pause: a thousand bets fit into a lunch break. Set deposit and loss limits in Stake Smart before the first bet, a bet count and a stop-loss in the Auto tab, and treat the 1% as the ticket price. Stake does not serve Ontario and is an offshore, Curaçao-licensed operator elsewhere in Canada. 18+. If the game stops being a game, our responsible gambling page lists tools and help lines.

Play Limbo on Stake with code PACTBONUS

CodePACTBONUS
Claim the deal

200% bonus + 3.5% rakeback · 18+ · Wagering terms apply

Common questions

What is the Limbo casino game?

A Stake Original, not the Playdead video game. You type a target multiplier between 1.01× and 1,000,000×, the game reveals a result multiplier, and if the result is at or above your target you are paid stake times target; otherwise the stake is lost. One input, one click, a result in under a second.

What is the RTP of Stake Limbo?

99.00%, shown in the game panel, at every target. The win chance is 0.99 divided by the target, so win chance times payout is always 0.99: $99 back per $100 wagered in the long run. The target changes only how bumpy the road is.

Is Limbo the same as Crash?

Same payout curve, different game. In both, the chance of reaching a multiplier m is 0.99 / m. Crash is multiplayer with a shared rising round and a cash-out you time; Limbo is single-player and instant, with the target fixed before the bet and no timing at all. The 1% house edge is identical.

Does a Limbo predictor work?

No. The next result depends on the server seed, which stays secret until you rotate the pair, and rotating starts a new pair. Predictor apps, Telegram signals and seed hacks are random guesses or phishing. Stake Limbo has no download; play it on stake.com.

What is the maximum multiplier on Stake Limbo?

1,000,000×. The target field rejects anything higher (“Maximum is 1000000”) and the Game Info rules state the same ceiling; the minimum target is 1.01×. A $1 bet at the ceiling pays $1,000,000 once in 1,010,101 rounds on average, with the win chance shown as 0.000099%.

What is the best Limbo strategy?

None beats the 1%. A low target such as 1.1× wins 90% of bets and loses the same $10 per $1,000 slowly; Martingale at 2× meets seven straight losses about once in 119 attempts and then needs a 128-unit bet. The useful choices are the stake, the target and a stop-loss set in the Auto tab before the first bet.

Does Limbo count toward the Stake welcome bonus, and can I play from Canada?

At the time of writing Stake Originals count 100% toward the 40× rollover, with a max bet of about 1% of the deposit; check the terms page. Stake accepts players in Canadian provinces and territories other than Ontario, where it is not registered with iGaming Ontario. 18+; our responsible gambling page lists limits and help lines.

Sources: stake.com/casino/games/limbo (game panel RTP 99.00%, tags Provably Fair and Stake Originals, the description with the 1,000,000x maximum win and the 1.01x minimum target, the target and win-chance fields with their validation messages, the default 2.00x / 49.50%, the Auto tab with the Advanced switch, the settings menu, the hotkeys panel, the Game Info rules and Max Betting Limits tabs, the Fairness modal with Overview, Seeds and Verify; opened 2026-10-11, English edition); stake.com/provably-fair/game-events (the Limbo rule: 0.99 / float, floored to two decimals, minimum 1.00; read 2026-10-11, served as the Finnish edition); stake.com/provably-fair/calculation via our provably fair guide (the 0.380899459356 float, Dice, nonce 1, 2026-10-05); our own arithmetic and a 2,000-session simulation for the EV, standard-deviation and losing-streak table and the 2.59x example; Stake bonus terms as published at the time of writing.