Knife Rush

Forty-five rounds of Knife Rush cost us $44.12 and paid back $37.51: 191 knives thrown, 174 stuck, 17 rounds lost. Knife Rush is a knife-throwing ladder game from InOut Games, released on 10 November 2026 with a published RTP of 95.5%. Every knife that sticks in a spinning target moves the cash-out one step up a printed ladder, and a knife that strikes another knife takes the whole stake. We played InOut’s own build on 11 November, which is free in a browser tab and wants no registration or deposit, read every rule screen, and then spent nine rounds throwing at the worst possible moment to find out whether aim counts for anything.

What is Knife Rush?

Knife Rush is a single-player crash game dressed as a knife-throwing act. A round log spins against a mountain landscape with a handful of knives and a couple of apples already stuck in its rim. You set a bet from $0.12 to $200, pick one of four difficulty levels and throw. The number painted on the log is the multiplier the next knife will lock in, and the yellow CASHOUT button shows what you can take right now.

Under the paint it is a step game. Knife Rush belongs with InOut titles like KO Million, where each landed punch is one rung of a multiplier ladder and a hidden knockout ends the fight; here the rung is a knife and the knockout is a clash of blades. There are no reels, no paylines and no bonus to buy. The only decision that changes your result is when to stop.

How does a round of Knife Rush work?

A Knife Rush round is a chain of single throws, and each one either raises the cash-out or ends the round. The sequence below is what happened in every round we played.

1. Bet and difficulty. Both lock when the round starts. Easy opens at 1.01x for the first knife, Hardcore at 1.44x.
2. Play throws the first knife. There is no free look. If that knife clashes, the stake is gone before the button has turned into GO.
3. GO throws the next one. A stuck knife moves the yellow marker one step along the ladder under the target and updates CASHOUT.
4. CASHOUT ends the round with a win. A green YOU WIN plate shows the amount, and the log resets for the next bet.
5. A clash ends it with nothing. The screen flashes dark red, the multiplier disappears and the round goes into the history as x0.00.

A full round crosses three targets, counted 1/3, 2/3 and 3/3 at the top of the screen. The wooden log comes first. When it has taken its share of knives it splits into pieces and a red and white bullseye drops in, and after the bullseye comes a wheel of cheese. On Easy each target takes ten knives. The first target took nine on Medium, eight on Hard and six on Hardcore.

45 rounds on four difficulty levels

We staked $1 a round in 44 rounds and $0.12 in the opening one, and took every figure in the table from the game’s My bet history panel. Cash-out points were fixed in advance: three, six and nine knives on Easy, three and five on Medium, two and four on Hard, one and two on Hardcore, four in the timing test described below, plus at least one round per level played until it broke.

Easy and Medium finished within a dollar of even, which is what short cash-outs on gentle ladders tend to do. Hard was the hole in the session. Three rounds banked 1.46x after two knives, and the other seven lost: four of them on the first knife, two on the second, and one after ten knives with $12.52 on the button and 17.74x painted on the bullseye. That round is the game in one picture. On Hardcore the two rounds we let run reached 15.30x and 5.53x before the clash.

The longest run of the day was the very first round, 22 knives on Easy at the $0.12 minimum. It broke the log, cleared the bullseye and put two knives into the cheese, the CASHOUT button stood at $0.39 for 3.28x, and knife 23 clashed on the way to 3.70x.

Does timing matter in Knife Rush?

No, not in any way we could measure. The rules tell you to time your throws, and the target does spin with gaps and blades passing the six o’clock point where your knife flies in. So we tested it. A script watched a small box right under the target and pressed the button only when a knife handle or an apple filled it, which in a real knife-throwing game is the one moment you must not throw.

Across nine Easy rounds, 34 knives were released that way and 33 stuck. The ladder odds for those same steps, worked out in the next sections, predict 32.8. The one aimed throw that failed was a second knife, a step that fails about once in fifty anyway. The knife on screen lands next to the blade it was pointed at, and the multiplier ticks up as if nothing had been in the way.

Two more details point the same direction. The autoplay panel takes a cash-out step from 1 to 30 and a run of 10 to 200 rounds, and it throws without waiting for a gap. And each round’s fairness card carries a server seed that was hashed before the round began. That makes Knife Rush the reverse of BalloniX, where the moment you let go is the entire decision. Here the moment you press is decoration, and the choice that counts is whether to press again.

The rulebook against 45 rounds

The How to play? window in Knife Rush has five numbered points and the Game rules window has three lines of limits. The left column quotes them; the right column is what the demo did when we played.

Two rows matter more than the rest. The timing line is flavour text, as the test above shows. And the first-knife risk is easy to miss: on Hardcore the opening throw alone should fail about one time in three, and in our nine Hardcore rounds it failed three times.

Four ladders: 30, 25, 22 and 18 knives

The whole ladder is printed under the target, five steps at a time, with arrows to scroll it before you bet. If the 95.5% return holds wherever you stop, the chance of reaching a step is 0.955 divided by its multiplier. The table applies that to the first knife, the knife that breaks the first target and the last knife of each level.

The same division gives the risk of any single throw: the multiplier you already hold divided by the one you are throwing for. On Easy, going from 1.01x to 1.03x succeeds about 98 times in 100. On Hardcore, going from 1.44x to 2.21x succeeds about 65 times in 100. Our 191 throws ended in 17 clashes where these odds predict about 21, well inside what luck does in a sample this small.

The ladders were not drawn for this game. Thirty steps to about 23x, 25 to about 2,457x, 22 to about 62,162x and 18 on Hardcore are the lane counts and ceilings of Rabbit Road, which also runs at 95.5%. Knife Rush shortens the big numbers to x2.46K, x62.2K and x3.61M. If you know how far you usually walk the rabbit, you already know how many knives to throw.

What is the max win in Knife Rush?

The maximum win in Knife Rush is $20,000 per round, and the Game rules window states it next to the $0.12 minimum and $200 maximum bet. The top of the Hardcore ladder reads x3.61M, but the cap wins that argument at every stake: even twelve cents times 3.61 million is $433,200. A full 18-knife Hardcore run can only be paid in full on a bet of about half a cent, which the game does not accept.

Provider	InOut Games
Game type	Knife-throw step ladder with cash-out
RTP	95.5% (official page)
Difficulty	Easy 30 knives, Medium 25, Hard 22, Hardcore 18
Targets per round	Three: log, bullseye, cheese wheel
Top step printed	23.24x Easy, x2.46K Medium, x62.2K Hard, x3.61M Hardcore
Bet limits	$0.12 to $200; max win $20,000 (demo rules panel)
Our session	45 rounds, $44.12 staked, $37.51 returned
Volatility	Not published; set by the difficulty level
Fairness	Provably fair; server seed check reproduced with SHA-512
Platforms	HTML5 in the browser
Released	10 November 2026

The cap bites lower down as well. At the $200 maximum it is reached at 100x, which on Hardcore is the tenth knife (185.08x) and on Hard the sixteenth (166.87x). The Hard ceiling of x62.2K is payable in full up to a $0.32 bet and the Medium ceiling of x2.46K up to about $8. Only Easy never touches it: 23.24x on $200 is $4,648.

Is Knife Rush fair? One round checked with SHA-512

Knife Rush is provably fair in the sense InOut uses for all its step games: the server seed behind a round is hashed and shown before the result exists, and revealed afterwards. We opened the card of our first round, the lost $0.12 one, from the shield icon in My bet history.

The check is one line of Python: python -c "import hashlib;print(hashlib.sha512(b'ce6a9883ac7fdf167a407fa3948b863f899012b4').hexdigest())" prints a 128-character string that starts 1c0473ca and ends c043eda8, the Hashed Server Seed on the card. The settings screen labels that value SHA256; its length and this result say SHA-512. The Decimal line is the combined hash read as one number, which we also reproduced.

What stays closed is the step between the seeds and the knife that clashes. The fairness screen says the result comes from the server seed and the first three bets of the round, and InOut does not publish the formula. We compared the seed with the hash on the card after the round; to rule out a swap yourself, copy the hashed seed from Provably fair settings before you press Play.

Apples, the Skin Shop and what they are worth

Apples in Knife Rush are a collectible with no cash value. They sit on the rim of the target between the knives, and the counter on the Shop tile had reached 23 after our 45 rounds. The shop holds eight knife designs. Fang, a plain steel blade marked Common, is equipped from the start and a second knife is already owned. The other six are priced in pairs: two at 25 apples, two at 50 and two at 100. The rules add that apples cannot be bought with real money, and the shop repeats that skins are cosmetic.

Knife Rush APK, hack and predictor downloads

InOut distributes Knife Rush as a browser game and has put out no installable file, so a “Knife Rush APK” or “Knife Rush hack” in a search result was built by someone else. A predictor has even less to work with here than in most games: our 34 badly aimed knives show that the picture on screen does not decide the throw, and the seed that does is only revealed when the round is over. The reasoning is the same one we set out for Chicken Road 2 predictors and signal bots.

Three knives or thirty: who Knife Rush suits

This is a known ladder in a better costume. The throw has weight, the log bursting into planks is a real reward for getting through ten knives, and the apple shop gives low-stake rounds something to collect. What it does not have is the skill its rulebook hints at, and a player who believes the timing line will blame their thumb for losses a seed decided. Our own numbers argue for modest targets: Easy and Medium, mostly cashed between three and nine knives, returned $25.83 on $25.12, while Hard and Hardcore returned $11.68 on $19.00. Play it as a stop-or-go game at 95.5% and it is one of the more likeable ways InOut has found to dress that question.

Short answers about Knife Rush