Stake Limbo verifier
This Stake Limbo verifier recomputes a multiplier from the revealed server seed, your client seed and the bet's nonce, in your browser, using the construction Stake publishes: HMAC-SHA256(server seed, "clientSeed:nonce:cursor") read as floats, one float for Limbo. Stake publishes 99% RTP for Limbo (checked 2026-09-23). With the example inputs below the multiplier comes out as 1.95×; press “Load example values” and Verify to reproduce it.
Disclosure: VerifyRoll is sponsored by Craze, a competitor of Stake. Figures here are computed or taken from what operators publish, dated; nothing is estimated.
Loading verifier…
How a Stake Limbo multiplier is derived
How a Stake-scheme Limbo result is computed from one float, why the 0.99 factor in the formula is the house edge, and how often a target is reached.
- Take the first float from HMAC-SHA256(server seed, "clientSeed:nonce:0").
- Compute 100,000,000 ÷ (float × 100,000,000 + 1). Small floats give large values; the +1 stops division by zero and caps the maximum at 100,000,000.
- Multiply by (1 − 0.01). This is where the edge enters: every raw multiplier is scaled down by 1%.
- Floor to two decimals and never go below 1.00.
multiplier = max(1, floor((1e8 / (float × 1e8 + 1)) × 0.99 × 100) / 100)
Full derivation, payout rule and every operator's figure on the Limbo maths page.
Stake's published RTP for Limbo
99%, a house edge of 1%, as published by Stake and checked 2026-09-23. The edge is the 0.99 factor inside the formula. Without it, the multiplier distribution would return exactly 100% at every target. The verifier proves the multiplier; it cannot see the RTP. Read the figure in the game's own info panel and compare it with this one.
Worked example
Computed on the server by the same function the verifier runs. The seed pair is the site's example, at nonce 7 on this page; it is not from a real Stake round.
- server seed
- 2b4d2f1a7c9e0f3b8a6d5c4e3f2a1b0c9d8e7f6a5b4c3d2e1f0a9b8c7d6e5f4a
- client seed
- verifyroll
- nonce
- 7
- HMAC-SHA256(server seed, “verifyroll:7:0”)81727d41db6b44706ddfd681ab9cd0124cc482392adbbdda8ac8ac626566bf1a
- Float (1): 0.50565322
- float = 0.5056532176677138
- raw = 1e8 / (float × 1e8 + 1) × (1 − 0.01) = 1.957864
- multiplier = max(1, floor(raw × 100) / 100) = 1.95×
Questions
- What do I paste to verify a Stake Limbo multiplier?
- The revealed server seed, your client seed and the bet's nonce, plus the server seed hash that was shown before play if you want the commitment checked. Limbo consumes one float from the HMAC output.
- What RTP does Stake publish for Limbo?
- 99%, a house edge of 1%, as published by Stake and checked 2026-09-23. The edge is the 0.99 factor inside the formula. Without it, the multiplier distribution would return exactly 100% at every target.
- Is Limbo the same maths as Crash?
- The reach probability is the same shape, 0.99 ÷ target, but the randomness comes from a different place: Limbo uses the per-bet HMAC float, Crash uses a pre-committed hash chain shared by all players in the round.
- Why does the formula add 1 to float × 1e8?
- So the denominator can never be zero when the float is 0, and so the multiplier has a hard ceiling of 100,000,000 rather than infinity.
Next
- Verify any Stake bet: where the seeds are, all seven games.
- Shuffle Limbo verifier: same construction, separate operator.
- Stake Plinko verifier
- Stake Mines verifier
- Chance of reaching a target (same 0.99/x distribution)
- Stake in the casino directory
last reviewed 2026-09-27 · derivation as implemented in @craze/fairness