Extended vs Aster in 2026

Breadth on one side, verifiability on the other

Aster lists 578 perpetual markets across six asset classes, the longest listing of any venue this desk reads (Aster docs, checked 18 September 2026). Extended lists 325 across five, including pre-IPO contracts (Extended docs, checked 18 September 2026). On raw coverage the gap is real and it favours Aster.

The gap reverses on what can be checked. Extended settles on Starknet, where every transaction is validated, and runs a public bug bounty of up to $500,000. Aster's own documentation states that core chain contracts and RPC infrastructure will not be open-sourced at the moment, and that external validator participation is not supported during phase one (Aster Chain overview, checked 18 September 2026). Its published audits cover vault, earn and stablecoin contracts rather than the chain that now matches trades.

Fees, leverage and the shape of a position

Aster charges 0% maker / 0.040% taker on its order book, so passive orders are free and a $10,000 taker round trip costs $8.00. Extended charges 0% maker / 0.025% taker for every account with no tiers, or $5.00 for the same round trip. For a maker-heavy strategy Aster is cheaper; for a taker-heavy one Extended is.

Leverage differs more than the fee. Aster's book reaches 200x on BTC but only up to 400 USDT of notional, stepping to 150x and then 100x as size grows; Extended holds 50x to $4 million of position value before stepping down. One is built for small aggressive positions, the other for size. At either end of that range the arithmetic underneath is the same. The margin behind a leveraged position is all it has, and a market that gaps can take the lot.

Where neither of them commits

Trading hours. Both list equities, commodities and currencies, and neither states one policy covering the lot. Extended comes closer: its real-world-asset pages put the order-book markets at 24/7 and warn that the Prime markets behind a quote request run 24/5 or only in the underlying session, halting while positions and margin requirements stay live. Aster says nothing about hours in the documentation we read on 18 September 2026. That leaves at least one side of this page guessing about the thing that decides whether a non-crypto position can be closed on a Sunday — the question EVEDEX answers by stating that all of its markets trade around the clock (18 September 2026, EVEDEX documentation), and Ostium answers by publishing each session in full.

Which one fits which trader

Aster suits a trader who wants the longest market list and trades mostly as a maker, and who accepts a chain whose contracts are closed and whose validation is not yet external. Extended suits one who wants audited settlement, a bug bounty and a flat taker rate, and can live with a shorter listing and a 50x ceiling on BTC.

How the comparison was made

Every column here but one is read rather than measured. On 18 September 2026 the desk read each venue's terms and documentation for what it lists, when those markets are open, what it charges, who can move the collateral. Depth nobody publishes, so the desk took it: each BTC perpetual book read on a timer, ten minutes apart, closing 17:58 UTC on 20 September 2026. The desk keeps an account at none of them. The same five criteria, settled on 23 September 2026 and set out on the method page, order the 2 venues scored here.

The hours are the softest column here, and one page is ordered by them — a line in a venue's own document about when a market is open, not something anyone watched. Custody is read the same way. Depth is measured, but only on BTC, the one contract all of these venues run; thickness there says nothing about the XAUT book beside it. Fees are the day-one rate, nothing a trader earns back taken off.