Ranked & Reviewed
Best Perpetual DEXs
On-chain perpetuals venues ranked on funding costs, liquidation mechanics, oracle design and depth under stress.
No money changes hands. This is a non-commercial project: no advertising, no sponsored placements, and no affiliate or referral links. Links to the services below earn us nothing. Every overall score is the plain average of the criteria shown on the page, so you can check the arithmetic yourself — see our rating methodology and how the site is funded.
Ostium publishes an opening fee of 3 to 10 basis points by asset, a flat $0.10 oracle fee refunded on full closes, a rollover fee in place of funding, and a liquidation formula backed by a stated 25% collateral backstop. Its public contracts are MIT and six audit reports are named.
Pros
- Opening fee published as 3-10 bps by asset; oracle fee a flat $0.10 USDC, refunded on full closes
- No separate closing or liquidation fee charged to the trader
- Liquidation threshold given as a published formula, with a 25% collateral backstop on all leveraged trades
- Six audit reports named: Zellic, ThreeSigma and Pashov Audit Group, dated 2024 to 2026
- Public contracts under the MIT licence
Cons
- Maker/taker structure is not specified; it uses opening fees instead
- Uses a rollover fee anchored to real-world carry rather than a conventional funding rate, which is less familiar
- Some repositories in the organisation carry no licence
dYdX v4 publishes entry-tier fees of 1.0 bps maker and 5.0 bps taker, an hourly funding rate with its formula, and a liquidation penalty capped by default at 1.5% flowing to an insurance fund that absorbs liquidation profit and loss. Its chain is licensed under AGPL-3.0.
Pros
- AGPL-3.0 on the v4-chain core — a strong copyleft open-source licence covering the actual protocol
- Entry-tier fees published: 1.0 bps maker, 5.0 bps taker
- Funding charged hourly with the formula and a 0% interest component for most markets published
- Liquidation penalty capped at 1.5% by default; insurance fund absorbs liquidation profit and loss
Cons
- No official audits page could be reached, so no auditing firm is named
- Maintenance margin varies by asset and must be read per market
GMX V2 charges a position fee of 0.04% or 0.06% depending on whether a trade reduces or increases open-interest imbalance, accrues funding continuously, and publishes a tiered liquidation fee from 0.20% to 0.45%. Its core is BUSL until 2026, audited by five named firms.
Pros
- Position fee published: 0.04% if it reduces imbalance, 0.06% if it increases it
- Funding accrues per second based on long/short imbalance
- Liquidation fee published by market type: 0.20%, 0.30% and 0.45%
- Five auditors named: Guardian, ABDK, Certora, Dedaub and Sherlock, with a bug bounty
Cons
- Business Source License 1.1 on the V2 core until the earlier of 31 August 2026 or four years
- No insurance fund is described; risk sits with the liquidity pool
- Minimum collateral thresholds vary by market
Drift's protocol-v2 repository is Apache-2.0 and audited by Trail of Bits and Neodyme. Its official documentation URL now redirects to Velocity, a self-described fork, so the fee and mechanism figures below are read from that redirected site and flagged as such.
Pros
- protocol-v2 core repository is Apache-2.0, a recognised open-source licence
- Audited by Trail of Bits and Neodyme, with reports linked and dated
- Funding, liquidation and insurance-fund waterfall documented in structure
- Entry-tier fees available on the served documentation: 0.035% taker, small maker rebate
Cons
- The official docs URL redirects to Velocity Protocol, a self-described Drift v2 fork, so figures are not read from a canonical Drift-branded page
- OtterSec audit is noted as in progress rather than complete
- Fee and mechanism figures should be treated with that redirect caveat
Hyperliquid publishes entry-tier fees of 0.015% maker and 0.045% taker, an hourly funding formula with its exact parameters, and states there is no clearance fee on liquidations — with a maintenance margin band of 1.25% to 16.7% by leverage. Its core matching engine is not open source.
Pros
- Entry-tier fees published: 0.015% maker, 0.045% taker
- Funding paid hourly with the full formula and its clamps published
- No clearance fee on liquidations, stated plainly
- Maintenance margin band published: 1.25% at 40x down to 16.7% at 3x
Cons
- The core matching engine is not open-sourced; only SDKs are MIT and the node is Apache-2.0
- The contracts repository carries no licence
- Only Zellic is named, for the bridge contract rather than the exchange core
Orderly publishes an entry-tier taker fee of 3.0 basis points with a zero maker rebate cap, releases its EVM contracts under Apache-2.0, and names Halborn, Zellic and Guardian Audits. Funding and liquidation terms were not verifiable on the pages read.
Pros
- Entry taker fee published at 3.0 bps for crypto and RWA markets
- Apache-2.0 on the contract-evm core repository
- Three auditors named — Halborn, Zellic, Guardian Audits — with reports in a dedicated repo
Cons
- Funding mechanism was not verified on the pages reached
- Liquidation terms were not verified on the pages reached
- The fee page is framed around builder tiers, complicating a plain retail read
Aevo publishes a standard taker fee of 0.08% and maker 0.05%, hourly funding with its formula, and a tiered liquidation fee from 1% to 5% with maintenance margins by market. No open-source core repository or named audit could be verified.
Pros
- Standard fees published: 0.08% taker, 0.05% maker, with tiers available
- Funding paid hourly with the premium-and-clamp formula published
- Liquidation fee published by tier: 1%, 3% and 5%
- Maintenance margin stated: 3% for ETH and BTC perps, 48% for pre-launch futures
Cons
- No open-source core contracts repository could be confirmed; the SDK licence endpoint returned 404
- No auditing firm is named on the pages reached
- The clearest documentation was found in an aggregated text dump rather than structured pages
Gains Network publishes opening and closing fees by asset class — 0.035% on BTC and ETH, 0.05% to 0.06% on other cryptos, down to 0.012% on major forex — plus a per-block borrowing fee formula and published liquidation thresholds. Its core contract licence and audits could not be confirmed.
Pros
- Fees published in detail by asset class, from 0.012% forex to 0.06% non-core crypto
- Fixed spreads published: 0.005% per side on BTC and ETH, 0% on most cryptos
- Borrowing fee given as an explicit per-block formula
- Liquidation thresholds published, ranging 63% to 90% by asset class and leverage
Cons
- The core gTrade contracts repository licence could not be confirmed; only SDK and tooling repos were readable
- No dedicated audits or security page was found
- Risk backstop and insurance arrangements are not documented on the fee page
Jupiter Perps publishes a 0.06% base fee to open and close, a price-impact fee on imbalanced open interest, and an hourly borrow fee in place of funding. Liquidation forfeits remaining collateral to the JLP pool. The core program is not published and no audit could be verified.
Pros
- Base fee published at 0.06% of trade size to open and close
- Uses an hourly borrow fee based on utilisation, documented as compounding hourly
- Swap fees published: 10 bps non-stables, 2 bps stables
- Liquidation consequence stated plainly: remaining collateral is forfeited to the JLP pool
Cons
- The core on-chain perpetuals program is not published as an official open-source repository
- No audit could be verified from an official page
- The specific maintenance-margin percentage is not stated on the fee page
Paradex publishes a base-tier maker fee of 0.3 basis points and taker of 0.045%, among the lowest maker rates here. Beyond fees, most of its documentation — funding, liquidation, audits and the core licence — could not be verified.
Pros
- Base-tier maker fee published at a low 0.3 bps (0.003%)
- Taker fee published at 0.045%
- The Python SDK is MIT-licensed
Cons
- The funding-rate page returned 404
- Liquidation terms could not be reached
- The audits page returned 404 and no firm is named
- Only the Python SDK licence is confirmed, not the core protocol
At a glance
| Service | Score | Best for |
|---|---|---|
| Ostium | 8.6 | MIT-licensed with six named audit reports |
| dYdX | 7.8 | AGPL-licensed core with a published liquidation cap |
| GMX | 7.8 | Imbalance-based fees, five named auditors |
| Drift | 7.2 | Apache-2.0 core; live docs now serve a fork |
| Hyperliquid | 7.2 | Hourly funding and liquidations documented in full |
| Orderly Network | 6.2 | Apache-2.0 core with three named auditors |
| Aevo | 5.4 | Fees and liquidation tiers published, code unverified |
| Gains Network | 5.2 | Granular fee schedule, core licence unconfirmed |
| Jupiter Perps | 4.6 | Fees documented, core program closed |
| Paradex | 2.8 | Low maker fee published, wider docs unreachable |
Perpetual futures have no expiry, so a funding rate keeps the contract tethered to spot. That rate, charged repeatedly, usually outweighs the trading fee for anyone holding a position for more than a few hours.
What matters most on a perpetuals venue is what happens when the market moves fast: how liquidations are triggered and priced, which oracle feeds the mark price, how large the insurance fund is, and whether profitable traders have ever had positions cut to cover losses elsewhere.
How we score this category
A perpetuals venue is judged on what happens to a position under stress, not on advertised leverage, so this rubric scores the published cost of trading, the funding mechanism, the liquidation terms and the code licence. Where a protocol's core matching engine is closed while only its SDK is open, we say so — an open SDK is not an open exchange.
Every service in this ranking is scored 0–10 against the same 5 criteria, each carrying equal weight (20% apiece). The overall score is their plain average, rounded to one decimal place, and each review prints the individual numbers and the arithmetic so you can reproduce the result yourself.
What each criterion covers
- Fee transparency20% of the score
- Whether maker and taker rates at the entry tier are published as figures a trader can look up before opening a position, rather than described only in general terms.
- Funding & liquidation clarity20% of the score
- Whether the funding mechanism, its interval and the liquidation penalty or maintenance margin are documented as numbers rather than left to be discovered on the position screen.
- Licence & code openness20% of the score
- The licence on the core protocol repository, read from the file — not the SDK. A recognised open-source core scores highest; a delayed-open licence mid-table; a closed core, or only an open SDK wrapping a closed engine, scores low.
- Audit disclosure20% of the score
- Whether audit reports are published with the firms named and, ideally, dated. An unnamed or unreachable audit record is scored as absent.
- Risk documentation20% of the score
- Whether the insurance or backstop mechanism, the maintenance-margin schedule and the treatment of bad debt are explained, so a trader can understand what stands behind the venue when a position blows through its collateral.
What does not affect a score
Nothing commercial, because there is nothing commercial to affect it. This is a non-commercial project: no advertising, no sponsored placements, and no affiliate or referral links anywhere on the site. Links to the services above earn us nothing, so no provider can buy a score, a rank, or an entry in this list. Where two services finish level, they are listed alphabetically rather than ordered silently.
The process across all categories, including how often we re-check, is set out in our ratings methodology.
Rankings are editorial. Nothing here is financial advice. Editorial policy · How we score · How we're funded.