Best Hyperliquid prop firms 2026 — every firm checked one by one, scored on a public grid

The short answer: Propr first, Hypernova second, HyperPNL third

Propr.xyz leads at 86.7/100 and Hypernova follows at 85.7 — the two we bought, traded and were paid by, 48 and 6 payouts reconciled on-chain. HyperPNL scores 64.5: bought and traded, no payout taken. GT Funded stays at 18.3 so the failure is visible. Grid reviewed 10 September 2026.

The short answer

In August 2026, Propr.xyz was the best Hyperliquid prop firm — the one with the longest operating track record and turnkey automation. It is open at every size with a published rulebook, six account sizes from $5K to $200K, an 80% profit split, USDC payouts inside 24 hours, and — decisive for me — explicit permission for bots, copy trading and API access. Hypernova is a serious challenger and went public on 14 August 2026, with published pricing and no bot restrictions in its public rules. HyperPNL sells two evaluations but caps at $25K and publishes written rules for only one of them. That is the whole verdict; the rest of this page is the evidence — and the grid of 10 September 2026 now puts the same three firms in the same order, by arithmetic rather than by opinion.

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

If what you actually want is the step-by-step route rather than the comparison, we wrote it separately: what a Hyperliquid funded account actually is walks through the five steps, tier by tier, with the current fee grid.

The ranking: funded firms on Hyperliquid, Trust Score descending

Every firm on this venue that puts real capital behind a trader sits in the table below, sorted by total, highest first. The total is the weighted sum of five criteria, and the Status column says whether we paid for an account or only read the documentation. The mirror platforms are further down: they are not funded accounts, so they are not ranked here.

Grid reviewed 10 September 2026 · statuses checked 1 October 2026

Looking wider than this one venue? best crypto prop firms 2026, all 7 ranked puts these firms next to the centralised ones and the other chains.

Firms that fund you with real capital

Three platforms on Hyperliquid put actual capital behind a trader who passes. They are not equally available: one sells evaluations at every size today, one is capped at $25K, and one cannot be bought at all yet.

Funded accounts vs mirror platforms — the distinction that decides everything

The distinction that decides everything

Before any split comparison makes sense, one question has to be answered: does passing lead to real capital, or to a simulation?

Both models are legitimate, and the simulated ones are often upfront about it in their legal text. The problem is that the marketing above the fold is identical — same challenge, same profit target, same “payout” — while the disclaimer sits in 8-point type in the footer. A trader comparing a 90% split against an 80% split without reading that footer is comparing a rewards programme against a funded account and concluding the wrong thing.

So this page sorts by model first and split second. Every figure below was read on the firm's own site, and where a firm does not publish a number, it says not published rather than borrowing one from a comparison table.

Why prop firms moved to Hyperliquid

A prop firm asks you to trust it twice: once that your fills are real, and once that it will pay you. Traditional firms answer both with a dashboard. Hyperliquid changes the answer to both.

Orders route to a public perpetuals order book, so your execution is the same book everyone else trades — a firm cannot quietly widen your spread or slip your stop, because it does not control the venue. And settlement can be written on-chain, which turns a payout from a claim into a transaction hash. That is the whole appeal, and it is why every serious on-chain firm launched here rather than building its own matching engine.

It also means the firm never holds your trading capital. You are not wiring money to an offshore entity and hoping — the evaluation fee is the only money that leaves your wallet, which is a structurally different risk from a traditional funded account.

The other half of the answer is procedural rather than conceptual: what a Hyperliquid funded account actually is goes from wallet to first USDC payout, with the cheapest entry at each firm.

Everything side by side

Propr vs Hypernova on Hyperliquid: API, payout rail, reserve, receipts

One point separates them on the grid of 10 September 2026, for opposite reasons. On automation Propr is the more permissive of the two — its comparison line reads “Allowed — open API” against Hypernova's “No bot restrictions (published)” — and the two payout rails differ in kind: USDC on-chain · median 12 min (Aug, self-reported) for Propr, Instant · smart contract for Hypernova. Hypernova pre-funds a payout reserve anyone can read on Arbitrum — $497,973.33 at our reading of 1 October 2026 — and publishes a 5.9-second average payout time (read on its own homepage, 19 September 2026), while Propr's distributor is funded per payout, so you can audit what has been paid but not what remains to pay with. Against that, both have paid us: 48 Propr receipts and 6 Hypernova withdrawals, each matched to the cent against a public ledger.

Head-to-head on what actually matters

Availability. Propr: open, buy a challenge today. Hypernova: public since 14 August 2026, buy on demand. HyperPNL: open and buyable, but $25K is the real ceiling — 50K and 100K were still marked Soon at our reading. Availability no longer settles the ranking — track record does.

Rules you can plan around. Propr publishes its rulebook with exact numbers for both formats; Hypernova's is public too (v1.1 at our read of 5 September 2026) and its tiers are clear. Propr's edge is choice — a static-drawdown 1-Step and a cheaper 2-Step — where Hypernova offers one format with tiered risk.

Cost. Propr's line reads From $25 · 25K at $275, Hypernova's From $25 · 5K now on the public configurator (21 Sep 2026) · 25K at $120 / $275 / $365. On the size both firms publish, the 25K, that is $275 against $120 / $275 / $365 depending on risk tier. Remember the fee is your entire risk — you never deposit trading capital on either firm.

Payouts. Both settle in USDC on-chain. Propr commits to a $20 minimum within 24 hours per its rulebook; Hypernova pays programmatically by smart contract, publishing a 5.9-second average (read 19 September 2026), 24/7, backed by an on-chain reserve read at $497,973.33 on 1 October 2026.

Automation. Propr allows bots, copy trading and API use explicitly — “Allowed — open API” — and has the turnkey tooling. Hypernova's line reads “No bot restrictions (published)”, but there is no ready-made executor. A semi-automated executor — where you choose the trade and the bot manages DCA entries, take-profit and stop-loss inside the rules — runs on Propr today.

Both fiches carry the full evidence: Propr.xyz review: fees, rules, 48 payouts verified and Hypernova review: traded, paid, reserve read on-chain.

HyperPNL: open, scored 64.5, and why

HyperPNL scores 64.5/100 on the same five criteria, which puts it third here and nowhere near the top of the wider table. It is not a paper firm: we bought a ticket, funded it and traded it. What holds the score down is the ceiling — its comparison line reads $25,000 (real ceiling, re-checked 21 Sep 2026) · 50K/100K “Soon” — the paperwork, and a payout rail we have never used: Announced on-chain (Treasury) · our own purchase was debited off-chain, no contract address published (9 Sep 2026).

HyperPNL: open, but sold ahead of its own paperwork

HyperPNL takes a different route: no token, no points, no airdrop to wait for. Its own docs describe a Payment Router that splits each challenge fee between a payout treasury, protocol revenue and a referral fee — who supplies the capital behind that treasury is not published. On 9 September 2026 we stopped reading and paid: $42 for a 1-Step Flex $5,000 evaluation, funded in USDC on Arbitrum and traded the same day, and that session corrects what we published on 5 September. It is not pre-production. The checkout sells two evaluations — 1-Step Flex at $42 / $86 / $215 and 2-Step Flex at $50 / $90 / $213 on the $5K / $10K / $25K sizes — with 50K and 100K still marked Soon and the slider locked, our eighth consecutive check at that date. There is no time limit on any phase. The 1-Step targets 10% against a 3% daily and a 5% static drawdown; the 2-Step targets 10% then 5% against a 5% daily and a 9% static drawdown, with 2 then 3 minimum profitable days. Those are two products, not the contradiction between homepage and docs we called them.

What keeps it third is the paperwork, and what we could not verify with our own money. The GitBook rulebook documents only the 2-Step, so whoever buys the 1-Step buys an evaluation whose checkout screen is the whole of the documentation. To its credit, the app states its limits as figures a buyer can check rather than percentages to interpret — a $4,850.00 daily floor and a $25 profitable-day trigger on a $5,000 account — and it requires no profitable day at all on the 1-Step, the counter reading 0 / 0. Correction to the morning of the same day: we had read that $25 trigger as a hidden requirement, and it is not one; it is the threshold at which a day would count. The bigger reservation is on-chain. We traced our own $42 purchase end to end on Arbitrum: the only on-chain event is a deposit into RelayDepository, the deposit contract of the third-party bridging protocol Relay, after which the $42 was debited off-chain in HyperPNL's internal ledger. No Payment Router transaction, no Payout Treasury allocation, no revenue split, no referral fee, and no published address for any of the three contracts the docs name. We are not saying those contracts do not exist — we are saying that for the part of the chain we control there is nothing to check, against a homepage that sells “Everyone can see everything on-chain”. The engine does enforce its rules: our $25,000 demo was closed automatically at 12:18 on 9 September with the cause named in the interface, Daily Loss Limit, at a final balance of $24,170.48 against $25,000.00. And the leverage is real: on the paid $5,000 account the app let us carry $24,594 of BTC/USD notional at 5x asset leverage, which turns the $150 daily budget into 0.61% of BTC movement and the $250 total drawdown into 1.02%. Execution cost measured on $60,720 of volume: about $33.59, roughly 0.055%, in line with the advertised 0.045% taker — a winning trade of +$6.94 still left the account $26.65 down for the day. Untested, and it is the whole question: no payout. The account is in Phase 1 and negative, so the 80% split and the “instant” claim remain open. The Terms of Use, reachable only once logged in, do name an operator: Hyperpnl Ltd., registration CO-425378, Grand Cayman, with Cayman law and binding Cayman arbitration under §12 — which corrects our 5 September claim that no entity, company or jurisdiction was published. The same Terms, §3.1, exclude 18 jurisdictions including the United States and its territories and Canada, against a homepage that advertises “No restrictions — Anyone can participate freely”.

If you want the cheapest ticket and a small account. HyperPNL advertises rules written into code and a model anyone can inspect on-chain; we bought a 1-Step Flex $5,000 evaluation for $42 on 9 September 2026, funded it in USDC on Arbitrum and traded it the same day, and the promise did not survive the purchase. Traced end to end, the $42 produced one on-chain event — a deposit into RelayDepository, the deposit contract of the third-party bridging protocol Relay — after which the fee was debited off-chain in HyperPNL's internal ledger, with no published address for the Payment Router, the Payout Treasury or the Data Recorder its docs name. There is, however, no time limit on any phase, no consistency rule, and no minimum profitable day on the 1-Step: the paid counter reads 0 / 0, which retires a criticism we published that same morning. It sells two evaluations, not one: 1-Step Flex at $42 / $86 / $215 and 2-Step Flex at $50 / $90 / $213 for the $5K / $10K / $25K sizes, both read on 9 September 2026 on app.hyperpnl.com — Start Challenge (login required). We got this wrong on 5 September too, when we called the differing drawdowns and phase counts a contradiction between homepage and docs: they are simply two different products, and the correction belongs here rather than in a footnote. The real trade-offs are a $25K ceiling with 50K and 100K still marked Soon, a GitBook rulebook that documents only the 2-Step, and Terms §3.1 — read the same day — excluding 18 jurisdictions including the United States and its territories and Canada.

Everything we paid for, rule by rule, is in the HyperPNL review.

Platforms that are not funded accounts, and the firms left out on purpose

Two categories share the label on this venue, and only one of them funds you. The first block below is the platform that mirrors your trades instead; the second is the names other lists put here and we do not, with the check for each.

Platforms that are not funded accounts

These are sold with the same vocabulary — challenge, profit target, drawdown, payout — and they advertise the highest splits in the category. They are a different product. The distinction below is not our opinion; it is taken from each platform's own disclaimer.

Listed elsewhere, deliberately left out here

Comparison sites in this niche copy each other, and three names keep appearing on “Hyperliquid prop firm” lists that do not survive a check. Here is what we found and why each is excluded — so you can disagree with the reasoning rather than the conclusion.

Which one to pick, by trader profile

Three profiles, three answers — and the same grid behind all three.

If you want real capital today, at any size. Propr.xyz is the only one of the three that sells evaluations across the full range and has a versioned rulebook you can read before paying. It is also the one I trade with my own money, which is why it is the only firm scored above 85 in our Trust Score — every other firm here is assessed from published documentation and capped accordingly.

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

If verifiable settlement matters most. Hypernova is the best-capitalised entrant and the only one with institutional backing — and since 14 August 2026 it is public: checkout open, pricing published — from $25 on the 5K account (in-app, 17 August 2026), $120/$275/$365 on the 25K — funding tiers from $5K to $200K, and even measured pass rates (24.6%/19.2%/27.9%, read 17 August 2026). Keep its own FAQ's nuance in mind: funded accounts remain simulated, and trader signals can feed its central book.

If you mainly want to monetise trading you already do. Hyperstack is worth a look precisely because it does not ask you to change platform — but go in knowing you are joining a rewards programme, not getting funded.

The automation angle nobody prices in

Challenges are rarely lost to the market; they're lost to the trader at 2am — the revenge trade, the doubled position, the “just this once” that clips the daily loss. That's why the rulebook clause about automation matters more than a percent of split: the firm that lets you encode your discipline is the firm you're most likely to pass. Semi-auto execution with hard guardrails on your own account, non-custodial, where the daily loss and the drawdown are hard-coded walls the bot never crosses, is legal on the firm at the top of this table. Use it.

FAQ, sources and method

Every balance quoted on this page was read from a public RPC, and the commands are published: the on-chain tier scale, every address we read lets you re-run each check yourself.

Further reading: Alex Firdaus's analysis of which Hyperliquid prop firms are actually on-chain — an outside read that reaches the same on-chain / off-chain split as our first-hand table above.

Go deeper

⚠️ Trading leveraged products carries a high risk of loss. Most crypto prop firms are unregulated offshore entities, and simulated platforms are not funded accounts. Every figure on this page was read on the firm's own site on 10 August 2026 and may have changed since — check before you pay. This page contains affiliate links to Propr.xyz: we earn a commission on sign-ups, at no cost to you and with no effect on the rankings, which follow our . Nothing here is investment advice.methodology.