Hypernova Review 2026 — traded, paid, payout reserve read on-chain

Hypernova is a Hyperliquid-based prop firm with $3M of pre-seed funding led by Lemniscap (reported by The Block, May 2026), an in-house terminal, and payouts settled by smart contract in USDC on Arbitrum. We are not writing this from its marketing pages: three Hypernova accounts have been traded for this review — one $25,000 assessment bought with our own money through the public checkout in late August 2026, and two provided to us by the firm — and nine payouts were taken between 28 and 31 August. We disclose that split because it matters, and we do not say which of the three produced those payouts, because we do not know. Access is open — the homepage still carried a "CLOSED BETA · INVITE ONLY" badge at our 5 September 2026 read, which is a display inconsistency on their side, not a gate on yours (see the hero capture below). Everything below is either something we did on those accounts or something we read on their published pages on 5 September 2026, with the date attached. We use a Hypernova referral link, in place since 17 August 2026.

Verdict: 85.7/100, traded, 6 withdrawals verified

Where Hypernova sits against everything else we have bought and traded is the other half of this page: best crypto prop firms 2026, all 7 ranked.

Our verdict

Hypernova does the hard part well. The rules are published, versioned and written into a contract on Arbitrum, the payout reserve is readable at a published address, and the app matched the rulebook to the dollar on a real account. And it pays: nine payouts on our own $25K funded account between 28 and 31 August 2026, the largest reconciled on-chain at exactly 80% to the cent. Declared conflict of interest: three Hypernova accounts have been traded for this review — one bought through the public checkout, two provided to us by the firm — and we do not say which of them those nine payouts came from, because we do not know. What holds it two points behind Propr on our grid of 4 October 2026 — 85.7 against 87.7 — is not integrity but tidiness: on 5 September 2026 two of its own pages priced the same account $275 and $280, the homepage advertised a $200K ceiling while the app's sign-in page said $300K, the "closed beta" badge was still up three weeks after we bought through the open checkout, and its documentation links rot. Add the two things its own FAQ says out loud — funded accounts are simulated, and trader signals may feed its central book — and you have a firm that is honest in its documents and untidy in its shop window. If you want the tighter fees on a small account, the $5K Tight assessment at $25 is the cheapest way to test the machine yourself. We use a Hypernova referral link, in place since 17 August 2026.

→ See the sub-scores on the Trust Score page

→ See where Hypernova lands in our tested ranking of the best Hyperliquid prop firms 2026.

→ Is Hypernova legit? 6 withdrawals reconciled on Arbitrum, the reserve, and how our funded account ended

Is Hypernova legit? The short answer

Access

Public checkout — we bought the $25K assessment through it in late August 2026. The homepage badge still read “CLOSED BETA · INVITE ONLY” at our 5 September 2026 read.

Rules

Written to a smart contract on Arbitrum — TradingAccounts (proxy), 0x429d…4fcb, 61 bytes of code, checked by us on 8 August 2026.

Reserve

$360,302.43 in USDC on Arbitrum, read 6 October 2026.

Paid us

6 withdrawals, 1,932.09 USDC, 29 August 2026 → 1 September 2026, every hash reconciled on 19 September 2026.

Our account

F2, $25K funded, breached on 2 September 2026 — daily drawdown (−$752.15 on a $24,510 start-of-day balance).

Roya Trust Score

85.7/100, second of 7 on our public grid of 4 October 2026.

Every one of those transactions is published with its hash, its block time and the amount it settled: the 6 receipts and the vault.

The payout reserve, read on-chain: $360,302.43 on 6 October 2026

This is the one number almost nobody in the category lets you check before you trade. Hypernova’s payout reserve is USDC sitting at two published Arbitrum addresses — an operational vault, 0x9209…4c67, and a cold reserve wallet, 0x43c5…da07. The vault address is the same one published as the firm’s identifier in this page’s structured data. Our own reading of the two balances on 6 October 2026 came to $360,302.43.

Hypernova payout reserve: $360,302.43 in USDC on Arbitrum, read 6 October 2026 (block 512245987) — every dated reading: Hypernova reserve tracker.

What that proves is capacity to pay: on the day of the reading, the money to settle withdrawals existed and anyone could see it. What it does not prove is revenue, profitability or solvency in any accounting sense — a reserve can be topped up from anywhere, and we have not traced where its inflows come from. It is also a moving number: it changes day to day, which is exactly why every figure on this page carries the date it was read rather than a claim of being current.

What the last week did to it

Between our chain read of 14 September 2026 and the one on 21 September 2026 (7 days), the reserve fell $217,415 — from $977,923 to $760,508. Over the same 7 days the firm’s own lifetime-payouts counter rose $202,881, from $822,890 to $1,025,771 (read 21 September 2026). That leaves roughly $14,534 of the drop unexplained by published payouts. We are reporting a correspondence, not an audit: we did not trace the individual transfers, and a reserve can be topped up or drawn down for reasons that have nothing to do with paying traders. The reserve has been re-read since — $360,302.43 on 6 October 2026, the line above — but the counter has not, so that later figure is not compared with it here.

The other gap worth knowing is the one between the chain and the counter. On 21 September 2026 the stats page displayed $774,482 for the same reserve we read at $760,508 on chain the same day — $13,974 more than the two addresses actually held. A display lag of that size is normal for a cached counter and we have seen it in both directions; it is a reason to read the addresses yourself rather than the widget, not a reason to distrust the firm.

The full series, one dated reading per line, with what each movement did and did not let us conclude: the Hypernova reserve tracker, every dated reading.

Your real odds, from their own dashboard: 23.5% of paying traders have ever been paid

Pass rate by account size

All-time, the firm’s own breakdown, read 21 September 2026 on hypernova.xyz/stats.

Two things to take from that table before the caveat. The $100K account passes most often (28.4% over 807 resolved assessments) and the $200K least (16.4%), which is the opposite of the intuition that smaller accounts are easier — the daily-loss limit is a percentage, so the room scales with the account while the market does not care how big yours is. And the $25K size, the one we bought, sits at 24.8% on only 153 resolutions, a sample small enough that it should move.

Where their numbers do not agree with each other

We publish this because a dashboard that contradicts itself is a thing a reader should know before paying, not because we think anything is being hidden: panels on a live dashboard often use different windows and filters and say so nowhere. Every line below names both sources and the date, so you can re-check it yourself.

• Pass rate, two counts on one page: the headline reads 22.8% (611 of 2,675 resolved) while the by-size breakdown directly below totals 22.6% (545 passed, 1,869 failed — 2,414 resolved). Neither panel states its window.
• Passes versus funded accounts: 611 assessments are counted as passed, and the all-time trader journey on the same page shows 294 funded accounts out of 803 paid. Nothing on the page reconciles the two.
• The $25K Low Risk fee is $280 in the rulebook fee schedule and $275 in the homepage configurator and its tier-comparison table — the same $5 gap we first recorded on 5 September 2026, still open sixteen days later.
• Leverage: the homepage configurator and its comparison table both print a flat “5x” cap, while the rulebook has said “per symbol, see Markets” since its v1.0.3 entry of 29 June 2026.
• Market count: the homepage says “Browse all 99 markets”, the rulebook's market-access table says 140 pairs (79 crypto + 61 TradFi).
• Minimum funding: the homepage headline offer starts at “$10k–$200k”, while the rulebook fee schedule and the configurator both start at $5,000.
• Pass rates again: the homepage tier table gives 19.3% Tight, 20.1% Low Risk, 32.2% Medium; the stats page gives neither set — its own tier chart showed 20.9% for Tight over 30 days at the same read.
• The homepage still carried a “closed beta · invite only” badge at this read, six weeks after we bought an assessment through its public checkout.

On the first of those, the arithmetic is ours: the headline counts 66 more passes on 261 more resolutions than the breakdown printed underneath it. Neither figure is wrong in a way we can demonstrate — they are simply two different populations with the same label, and the page never says which one answers “what are my chances”. Our own rule for figures like these is on how we test prop firms: a self-reported number is published as self-reported, dated, and never converted into a score.

Fees and tiers: the full rulebook v1.1 schedule, re-read 21 September 2026

Prices read on 5 September 2026 · hypernova.xyz/rulebook

Read in the rulebook v1.1 fee schedule on 5 September 2026. One discrepancy to know before paying: the rulebook prices the $25K Low Risk account at $280, while the homepage configurator showed $275 for the same account the same day — a $5 gap between two of the firm's own pages, so screenshot the checkout total you actually pay. A fourth tier, High risk, exists in the rulebook but is marked restricted and its fees are TBD; the $200K Medium fee is TBD too. See the fee-schedule and configurator captures below.

The full schedule, all six sizes, re-read 21 September 2026

The table above is our 5 September reading, four sizes deep. We went back through the whole fee schedule on 21 September 2026 and not one published fee had moved a dollar: the rulebook is still v1.1, its changelog’s newest entry is still 4 August 2026 (“Tight risk account added”, commit a37c87b), and the two missing prices are still missing. Here is the complete version, including the $10K and $50K sizes this page had never published.

Fees are one-time, non-refundable once the account is activated, and paid in USDC. The $5 gap between the rulebook’s $280 for the $25K Low Risk account and the $275 the homepage configurator was still showing on 21 September 2026 has now been open for sixteen days — screenshot the checkout total you actually pay. Compare these against everything else we have bought: the full fee comparator.

What you are actually trading: Hyperliquid prices, Hypernova's own engine

Three consequences that change what you should expect. First, slippage and fills are Hypernova’s, not a public order book’s — you cannot verify an individual fill the way you can verify a payout hash, and the on-chain part of this firm is the money and the rules, not the trades. Second, trading fees mirror Hyperliquid’s maker/taker schedule with no markup (section 16, same read), with no monthly, inactivity, data, platform or withdrawal fee — so the assessment fee really is the whole cost. Third, the firm’s FAQ says it may use trader-generated signals to trade its own central book, and that you are notified in-app if yours are used; that is a business model worth knowing about, and one we have no way to observe from outside.

On the catalogue, the rulebook’s market-access table splits 79 crypto perpetuals from 61 TradFi ones (equities, ETFs, commodities, indices, FX) for 140 combined, and you pick your package at checkout — while the homepage was still inviting readers to “browse all 99 markets” on the same day. Leverage is capped per symbol in the rulebook and shown as a flat 5x on the homepage configurator. Neither gap costs you money; both are reasons to read the rulebook rather than the landing page.

Two limits are easy to trip over if you plan to run more than one account. You may hold several funded accounts at once, but only up to $200,000 of aggregate funded balance, and only one active assessment per asset class at a time (section 15). Hedging the same instrument across two Hypernova accounts — or across a Hypernova account and one at another firm — is the first prohibited conduct in the list and ends both. One genuinely trader-friendly mechanic in the other direction: if your equity touches the profit target while positions are open, Hypernova closes them for you to realise the gain and pass you (section 8).

The published addresses behind all of this, which anyone can open: the accounts contract at 0x429D…4Fcb and the assessment-fee treasury — where the money you pay actually lands — at 0x924e…8240, both listed on the firm’s own stats page at our 21 September 2026 read. How we grade that kind of disclosure, firm by firm: our on-chain prop firm index.

The rule that ends accounts: daily drawdown — our $25K was breached on 2 September 2026

Two limits end the account, and the daily one is the trap. Max daily loss is a percentage of the PREVIOUS day's 00:00 UTC closing BALANCE — 3% Tight, 3% Low, 4% Medium, 5% High — but it is enforced on your EQUITY, with no grace period: an open position that dips below the line breaches you even if it would have come back. Second, the max drawdown is a static floor computed once from the starting balance (3% Tight, 6% Low, 7% Medium, 8% High) and it never moves, so the cushion does not grow with your profits. On a $25K Low Risk account read on 5 September 2026 that is $750 of daily room against yesterday's closing balance and a hard floor at $23,500. Holding over the weekend and trading news do not end the account — section 13 allows both, along with scalping, grid and every other style. Strategy switching is the one that reads as allowed and is not: at our 21 September 2026 re-read, section 14.7 of the prohibited conduct bans passing the assessment on one approach and then trading a materially different one on the funded account, and that route ends in termination without payout, not in a breach.

Our own account says the same thing in numbers. F2 was a $25K funded account: 7 trading days, 40 trades, a 88% win rate and 13.34% ROI, with $3,333.01 of cumulative P&L — and it still died on 2 September 2026, on the daily drawdown (−$752.15 on a $24,510 start-of-day balance). A win rate near nine in ten does not protect you from one bad session against a limit measured on equity. The 6 withdrawals we had already taken stayed paid.

Payouts we received and how we verified them

Method and threshold — 80% profit split, on demand, no minimum amount and no request limit, settled by smart contract in USDC on Arbitrum (rulebook v1.1, read 5 September 2026).

The dashboard listed nine payout requests over 28–31 August 2026 (read 5 September 2026); the 6 transactions above are the ones we could reconcile on Arbitrum — status OK, a USDC transfer to our own wallet, block time matching the dashboard to the minute. Every payout figure we publish is labelled by how we know it on all payouts we verified on-chain.

Screenshots, dated

Proof — what the firm's site showed us

Hypernova vs Propr: reserve, speed, API, receipts

Hypernova’s edge is what you can check before you trade. Its payout reserve sits at two published Arbitrum addresses and read $360,302.43 on 6 October 2026. Its withdrawals are fast in a way that is measurable rather than advertised: the firm’s own counters reported an average payout time of 5.9s, a minimum of 2s and a maximum of 14.6s when we read them on 7 September 2026 — their figure, not our measurement — and the 6 withdrawals we took on our funded account settled in USDC on Arbitrum, each with a public hash.

Propr answers the half Hypernova cannot. Its payouts run through a verified distributor contract on Ethereum, one transaction per payout; the contract itself holds ~no balance and is funded payout by payout, so it is not a pre-funded vault like Hypernova’s. Its reserve sits in three other wallets: Since 3 October 2026 Propr publishes a Proof of Reserve across three public wallets (cold and hot reserve on Ethereum, margin on Hyperliquid); we read them ourselves on 4 October 2026 at 02:43 UTC: $1,025,662.96 USDC of stablecoin, plus $76,715.38 of perps account value under open hedges. On top of that it has automation and depth — the comparator line for its bot policy reads “Allowed — open API”, against “No bot restrictions (published)” for Hypernova, whose rulebook permits your own automation on a single account while banning copy trading and signal services outright. And it has the longer ledger with us: 48 payouts received against 6 withdrawals here.

The automation gap is sharper than the comparator line makes it look, and it has a date. Propr’s REST API is live and documented; Hypernova’s developers section still carried the label “coming soon” over its typed REST + WebSocket and its TypeScript and Python SDKs when we read it on 21 September 2026. So today you can run a bot on Hypernova — the rulebook allows it and there is no restriction — but you run it against the terminal, not against an API the firm supports. If your edge is an agent rather than a chart, that is the whole decision: prop firms that allow bots and publish an API, tested.

On our grid of 4 October 2026 the two are two points apart — Propr 87.7, Hypernova 85.7 — and the gap sits in the rules and freedom criteria, not in payouts, where Hypernova scores higher than anything else we track. On receipts the count is 48 to 6, which measures how long we have traded each firm with our own money, not which one pays better: both paid, both reconciled to the cent. The full fee-by-fee, rule-by-rule side of it is in Propr.xyz review: fees, rules, 48 payouts verified.

FAQ, sources and method

KYC & excluded countries

Not stated on the public rulebook at our 5 September 2026 read. The rulebook v1.1 does not describe an identity-verification step, does not name a document set, and publishes no restricted-countries section — silence, which is not the same as "no KYC". We are not documenting an absence we did not test, so treat identity checks as unknown and ask Hypernova in writing before you buy if your jurisdiction or your privacy is the deciding factor.

One nuance our 21 September 2026 re-read adds: silence on KYC is not the same as collecting nothing. Section 14.5 of the prohibited conduct makes “providing false identity information at registration” a terminating offence, which means identity information is collected at registration even though no verification step, document list or restricted-country section is published anywhere in rulebook v1.1. If your jurisdiction or your privacy is the deciding factor, get it in writing before you pay.

Bots, API, copy trading, weekends

Your own bots, yes — on a single account. Rulebook v1.1 read 5 September 2026: "Automated strategies built and operated by the trader themselves on a single account are permitted." Section 2 of the prohibited conduct is the other half, and it is a closing offence: copy trading, signal services, following external signals, and off-the-shelf strategies marketed to pass evaluations are all banned. So automate your own edge, never someone else's feed, and do not mirror between accounts. Around that: no time limit, no minimum trading days, no consistency rule, news trading and weekend holding allowed. See the prohibited-conduct capture below.

Sources and method

Official site: hypernova.xyz · unaffiliated link. Fees, tiers, targets, drawdown, payout terms, prohibited conduct and the bot policy: the rulebook v1.1 at hypernova.xyz/rulebook, read 5 September 2026 and re-read in full on 21 September 2026 — version, changelog and fee schedule plus sections 8, 12, 13, 14, 15 and 16, from which this page’s complete fee table, the routing sentence, the account limits and the strategy-switching prohibition are taken. The funnel, the pass rates, the lifespan figures, the payout counters, the displayed reserve and the published contract addresses: the firm’s own hypernova.xyz/stats and homepage, read 21 September 2026 — their figures, labelled as theirs, none of them audited by us. Payout reserve: direct Arbitrum balance reads of the vault and the cold reserve wallet, latest 6 October 2026, full dated series on the reserve tracker. Contracts and addresses: checked by us with eth_getCode and USDC balanceOf, 8 August 2026, published on our on-chain prop firm index. Our own record — the $25K funded account F2, its 40 trades and its breach, and the 6 withdrawals totalling 1,932.09 USDC — was read on our own Hypernova account on 19 September 2026 and every withdrawal hash reconciled the same day against a public Arbitrum node. Trust Score and sub-scores: our public grid of 4 October 2026, method in how we test prop firms. No figure on this page comes from a testimonial or a press release.

Affiliate link. We earn a commission; it never changes what you pay — it is what funds your cashback.

Payout calculator: the gross profit that recovers your challenge fees →

See also: Propr vs Hypernova, head to head · Hypernova discount code: no code, but 5% cashback in USDC · Hypernova alternatives: 3 firms compared

Affiliate disclosure: we may earn a commission if you sign up to Propr.xyz, Hypernova or Carrot Funding via our links, at no cost to you. Specs may change — verify on each firm’s site. Informational only, not investment advice. Leveraged trading carries a high risk of loss.

Not investment advice. A challenge fee is money you can lose in full.