The stats, with receipts
Every figure on this page carries its exact measurement window and the file it came from — the audited replay pipeline for the flagship year, the September 2026 studies for everything since. No projections, no cherry-picks — and the parts we haven't finished validating say so out loud.
Member results — real accounts, shared anonymously
Screens our members shared in September 2026, from their own Topstep app or from the Kadosh dashboard. Shown with permission and without names, licence keys or account numbers. These are not part of the audit below: results vary by account, size and day, the red days are here too, and none of it promises future results. We don't post members' trade charts, because they show the bot's order levels.
Balance on the funded account after 2 days with the Enterprise Bot. +$1,049.26 of it was realized on 2026-09-16, and the Kadosh dashboard shows the same +$1,049.26 for this member that day.
Realized in one day on a 100K Combine (evaluation), 11 trades on v1.9.59, early September 2026. This member ran their own bigger sizes, 1 to 3 contracts per bot and full-size NQ, above what the app measured safe. Bigger size swings both ways.
Three trading days, Sep 14-16, one red day included (−$39.70). Kadosh dashboard, app-reported; the account type isn't shown on this screen.
Sep 14-16, three green days. Kadosh dashboard, app-reported; the account type isn't shown on this screen.
On 2026-09-18, plus the days before it: the newer app versions on the left, older ones on the right, red days (−$48.68, −$599.06, −$33.66) included. Account type not shown.
Posted 2026-09-02: 8 member apps reporting that day. This adds up every account the members run, including practice accounts, so not all of it is real money.
“I’m up $360. I went to sleep down $150.” — a member, on Discord
Audited full-year performance — flagship gold program
Replay window 2025-03-17 through 2026-03-13, full production guard chain, broker-true costs, adverse slippage charged on BOTH sides of every trade, fingerprinted data. Re-audited 2026-08-10: an earlier engine version under-charged slippage, and we corrected every number on this page downward rather than let a kind assumption stand.
| Configuration | Net (year) | Trades | Days | Worst dip | Violations |
|---|---|---|---|---|---|
| 1 contract · 50K account | +$3,844 | 494 | 230 | $816 | 0 |
| 2 contracts · 100K account | +$6,964 | 491 | 228 | $1,632 / $3,000 allowed | 0 |
| 3 contracts · 150K account | +$11,086 | 492 | 228 | $2,448 / $4,500 allowed | 0 |
Since the audited year: what got better, measured (updated 2026-09-14)
Every row below is a replay of the shipped engine on exchange bars with research fills (one contract or one micro, commissions charged), over the longest window that instrument has at one-minute resolution. Gold bars before July 8, 2026 exist in no source, so the gold windows are 37 sessions. Green sessions = sessions that ended positive. Deepest dip = the largest fall from a high point, measured on closed sessions. None of these are live fills unless the row says so.
| Bot and setting (as shipped) | Window | Net | Trades / win rate | Green sessions | Deepest dip | Before this upgrade / context |
|---|---|---|---|---|---|---|
| Enterprise Bot · gold program, current setting (since v1.9.64) | Jul 8 – Sep 11, 2026 (37) | +$1,860 | 52 / 78.9% (profit factor 2.3) | 17 of 37 | $160 | +$1,876 with a $432 dip and 15 green — same money, dip cut by 63% |
| Playbook Bot · gold card, current setting (since v1.9.64) | Jul 8 – Sep 11, 2026 (37) | +$1,724 | 50 / 78.0% | 17 of 37 | $224 | +$299, 47% win rate, 7 green, $515 dip — 5.8x the money, win rate up 31 points |
| Playbook Bot · Nasdaq sweeps on 1 MNQ micro, shipped card | 2026 through Sep 11 (174) | +$7,140 | 39% (winners about twice the losers) | 64.5% | $570 | before: the default ran full-size NQ, which a 50K account cannot hold through its trailing line. On micros it is positive in 2025 as well and never touched a loss line in either year |
| Playbook Bot · Nasdaq sweeps on 1 MNQ micro, last quarter | Jun 16 – Sep 11, 2026 (61) | +$2,820 per micro (+$5,640 at 2) | 352 / 39.2% | 33 of 61 | $407 (worst day $161) | the same card and the same edge as the row above, on the most recent quarter |
| London Gold Bot · LG-2 range breakout, 1 MGC (new 2026-09-13; first live session Sep 14) | Jun 16 – Sep 11, 2026 (56) | +$2,158 | 32 / 90.6% (29 winners) | 29 of 56 (24 sessions had no trade) | $475 | the classic London engine: −$463, 42% win rate, $1,073 dip over 36 sessions — replaced |
| Enterprise Bot · Nasdaq leg, “Protected” preset (wide stop, breakeven then trail) on micros | Jul 9 – Sep 10, 2026 (63 days), 5 MNQ micros on a 150K | +$5,408 at 1 tick, +$2,238 at 4 ticks | 53 / 98.1% (2 losers) | 96.3% | $1,120 | the legacy NQ exits: −$974 on the same 158 signals — the only NQ family that survived the hostile suite |
| Wolfpack · six presets, 1 MNQ per leg, scoreboard after modelled costs (not live fills) | Mar 26 – Sep 21, 2026 (127) | TOP2 +$18,566 · TOP4 +$16,324 · TOP1 +$13,254 · TOP5 +$12,460 · TOP3 +$12,295 · All-Months-Green +$3,920 | leg win rates 65–82% | all six presets green; All-Months-Green and TOP2 green 7 of 7 months | $729.72 (All-Months-Green) | shadow-only until v1.9.45; now places real orders on your switch — see the live-fills row below |
| Wolfpack · live fills on customer evaluation and funded accounts | Sep 8 – 21, 2026 (10 trading days) | +$2,364.10 | 137 legs / 80.3% | — | — | real broker fills, practice accounts excluded; live and model agreed on direction 36 of 36 (Sep 4–16) |
| NY AM Bot · shipped setting, forward shadow ledger, 1 MNQ micro | May 27 – Sep 17, 2026 (66) | +$2,462.27 per micro (+$25,263.50 per NQ) | 66 / 72.7% | 48 of 66 | $1,045.08 per micro | last 30 days: 12 of 16 won, +$739.81 per micro, profit factor 2.28 |
| Asia Reversal Bot · shipped setting, forward shadow ledger, 1 MNQ micro | May 27 – Sep 20, 2026 (82) | +$2,619.90 per micro (+$27,222.50 per NQ) | 115 / 48.7% | 56 of 82 | $380.87 per micro | average win about twice the average loss; profit factor 1.85 |
| London Gold Bot · LG-2, full record, 1 MGC | Mar 17, 2025 – Sep 15, 2026 (316) | +$4,801.88 | 186 / 74.7% | 13 of 17 months | $1,816.08 | last 30 trades 27 of 30 (90%); 0 of 166 start dates breached at 1 contract |
| Enterprise Bot · Nasdaq leg, “Protected”, full window, 1 MNQ micro | Mar 1 – Sep 10, 2026 (76 trading days) | +$2,237.20 per micro | 145 / 95.2% | 69 of 76 days | — | 7 of 7 months green |
| One-click setup (v1.9.70), one unit each | Window | Net | Green sessions | Deepest dip | What it switches on |
|---|---|---|---|---|---|
| Safe and Green (50K) | Jun 16 – Sep 11, 2026 (64) | +$9,253 | 70.3% | $614 | Playbook 1 micro, Wolfpack All-Months-Green, London LG-2, Enterprise gold |
| Balanced (50K) | same 64 | +$13,936 | 73.4% | $1,348 | Playbook 2 micros, Wolfpack TOP3, London LG-2, Enterprise gold |
| More Money (100K / 150K) | same 64 | +$16,755 | 70.3% | $1,378 | Playbook 3 micros, Wolfpack TOP3, London LG-2, Enterprise gold |
| Heaven on Earth preset (v1.10.1 – v1.10.3), 50K, 1x contracts | Window | Net | Days green | Winning trades | Topstep 50K pass (graded) | What it switches on |
|---|---|---|---|---|---|---|
| Green-day lock | May 27 – Sep 25, 2026 (88) | +$5,400 | 79 of 88 | 69% | 98.9% | Asia 2 MNQ, NY AM 1 MNQ Mon–Thu, Wolfpack All-Months-Green 1 MNQ per leg, taking turns by the hour |
| Steady 80% | same 88 | +$11,201 | 80 of 88 | 81% | 99.8% | Asia AM 4 MNQ, NY AM 1 MNQ Mon–Fri |
| Steady 87% | same 88 | +$11,144 | 80 of 88 | 87% | 99.3% | Asia AM 4 MNQ, NY AM 2 MNQ Mon–Fri |
| Aggressive 65% | same 88 | +$12,923 | 77 of 88 | 65% | 99.6% | Asia 2 MNQ, NY AM 1 MNQ Mon–Thu, Moore IB 4 MNQ (its two setups can be open together: up to 8 MNQ) |
| Aggressive+ | May 27 – Sep 25, 2026 (84) | +$20,996 | 80 of 84 | 70% | 99.3% | Asia AM 4 MNQ, NY AM 3 MNQ Mon–Thu, Moore IB 3 MNQ (up to 6 MNQ) |
Source files: the 2026-09-13 continuation report and studies B8 (two-year account walk), B10 (portfolio), the LG-2 router replay (32 of 32 trades reproduced to the cent by the live router), the Enterprise NQ mega search and hostile suite of 2026-09-10 (SACRED-ENTERPRISE-CONFIG/14-mega-search-90d) and the Wolfpack forward scoreboard generated 2026-09-10. The flagship gold program’s full audited year above is unchanged. Validation pass 2026-09-14: every figure on this page was re-read from its source file, the prop-firm rules on the product page were re-checked against Topstep’s and Lucid’s own help pages, and the release count comes from the live update feed. Update 2026-10-05: version, bot count and test count brought to v1.10.3, the Heaven on Earth presets table added from the app’s own preset texts, and Topstep’s payout, Express Funded, Combine and Max Loss pages re-read the same day (unchanged).
Monte Carlo — 10,000 alternate years
Daily results block-resampled 10,000 times (5-day blocks preserve streaky weeks), entirely inside the 2025-2026 validation window. Distribution below: 150K · 3 contracts.
| Robustness check | 1 ct / 50K | 3 ct / 150K |
|---|---|---|
| Profitable resampled years | 85.2% | 84.5% |
| Year with its 5 best days deleted | +$839 | +$2,071 |
| Emergency brake trip rate | 7–18% | 7–18% |
The Scaling Lab — build your road past $10K/month
Two kinds of unit. Audited unit = one 150K (or 50K) account running the flagship gold program alone, priced from the audited year and its Monte Carlo. Stack unit = one 50K account running the Balanced one-click setup (four bots), priced from its 64-session research ledger, Jun 16 – Sep 11, 2026 — three months, not yet a year. Copy tools (TradeCopia-class) mirror the master account's orders to additional funded accounts — Topstep takes the trade, the copier repeats it elsewhere. Drag the slider; the math updates from the measured distributions.
Source: audited year +$11,086 and Monte Carlo quartiles (median +$11,027), 150K · 3 contracts, window 2025-03-17..2026-03-13. Gross of firm profit splits and payout policies.
Want every bot, config, instrument and size at once, with the payout cycle and an equity chart? Build your stack on the product page →
The safe way to scale
- Master account on Safe mode, guards at firm defaults — the exact audited configuration.
- Add ONE mirror at a time; run it two payout cycles before adding the next.
- Identical configuration on every mirror — the copier repeats orders, the seatbelt caps size per account.
- Verify each firm's copy-trading and payout rules yourself before mirroring into it. Firms differ and rules change.
- Take payouts on schedule — banked payouts are the only money the market can't take back.
The aggressive way (eyes open)
- Aggressive risk mode raises bet size inside the validated caps — deeper swings, same seatbelt.
- Stack mirrors faster — accepting that a drawdown hits every account simultaneously.
- The brake and drawdown guards stay armed on every account. They are not optional at any risk mode.
- Expect guard-heavy months well below median. Roughly 1 account-year in 7 ends negative; at fleet scale you WILL see losing account-years.
- Never disable the daily lockout to "catch up." That is how accounts die.
Kadosh Bot Enterprise · All figures from the audited release pipeline with measurement windows as labeled. Futures trading involves substantial risk of loss and is not suitable for every investor. Historical replay is not a promise of future results. Prop-firm rules, splits and payout policies govern real outcomes. Nothing on this page is financial advice. Strategy logic is proprietary — sealed, encrypted, never disclosed. · Risk Disclosure · Terms · Privacy