The Complete Guide to Onchain Options

Perps already won leveraged directional trading onchain. Options are the category sitting right next to it that still hasn’t been decided — real demand exists, one venue currently dominates the numbers, and the whole market is still a small fraction the size of Deribit. This guide is the map: what an option actually is, why the first wave of onchain options protocols stalled out, who has a real book in 2026, and how to read the numbers you’ll see thrown around without getting misled by them.

Onchain options are still a small, concentrated market as of late 2026 — one venue (Derive) is credited in public research with the large majority of onchain premium volume, and the entire category remains a small fraction of what trades on Deribit. Read the rest of this guide with that concentration in mind: it’s the single most important fact about the category’s current risk shape.

Options in one page — no Greeks required

An option is a contract that gives the buyer the right, not the obligation, to buy (a call) or sell (a put) an asset at a fixed strike price, usually before or at a set expiry. You pay a premium for that right — that’s the number people get stuck on in explainer threads. It’s the price of the contract, not the strike price, and not a guaranteed profit.

	Perps	Options
Expiry	None	Usually fixed (some newer designs stream premium instead)
Cost to hold	Funding, paid continuously	Premium paid upfront (or streaming on some designs)
Buyer’s max loss	Liquidation — can exceed posted collateral	Premium paid, generally the ceiling
Writer/seller’s risk	Symmetric to the buyer’s	Can be large; an uncovered short call is theoretically open-ended
What you’re trading	Direction + leverage	Direction + time + implied volatility

A framing that’s shown up repeatedly in crypto-Twitter discussion of options: they sit somewhere between perps and prediction markets. You still have a strike and a defined payoff shape like a prediction market’s binary outcome, but you also carry time decay and implied-volatility pricing the way a perp trader never has to think about. That’s not a marketing line — it’s the actual reason options UX is harder to teach than a perp’s “long or short, pick your leverage.”

Payouts are zero-sum, the same structural family as perps. If a buyer gets paid, a writer — an individual, a pool of LPs, or a vault that warehoused the risk — funded it. A platform is a matching engine and a risk engine, not a source of yield out of nowhere. Any “income” or “yield” product built on options (see Rysk below) is compensating depositors for taking on that writer-side risk, not generating a return independent of it.

Why the first wave of onchain options protocols stalled (2020-2023)

Earlier names — Opyn, Hegic, Ribbon-style DeFi Option Vaults (DOVs), the original AMM-based Lyra — ran into a similar wall from different angles: high L1 gas made rolling weekly options expensive and annoying, AMM-based pricing got picked off by better-informed counterparties, market makers weren’t willing to warehouse real tail risk on thin books, and retail already had a simpler leverage product (perps) that scratched the same itch without requiring anyone to learn implied volatility.

What’s different by 2026, per public research and venue-level changes: dedicated L2s and appchains cut the cost of running a real multi-leg book; CLOB and RFQ models (Derive, Paradex/Paradigm) replaced a lot of pure-AMM pricing that used to get arbitraged against LPs; prediction markets taught a much broader audience how to think about a binary or defined payoff; and institutional options demand became visibly real — Coinbase’s acquisition of Deribit and CME launching 24/7 crypto options in 2026 are both signals that options demand in crypto is not purely a retail-speculation story. Onchain options premium volume reportedly printed cycle highs through 2026 — still on a much smaller base than perps or Deribit, which matters for how you read any of the numbers below.

How to read the numbers without getting misled

Every venue review and every “options are pumping” thread will throw around a few different numbers. They are not interchangeable:

• Premium volume — dollars actually paid for contracts. The best single “is anyone genuinely using this” metric for an options venue.
• Notional — the face value the contracts control, which is naturally much larger than premium volume. Don’t compare it 1:1 against perp trading volume; it measures something structurally different.
• Open interest (OI) — contracts still outstanding, i.e. live risk still on the books.
• TVL — means a different thing depending on the venue’s model. On a CLOB (Derive, Paradex), TVL is roughly trader margin. On an AMM or peer-to-pool design (Panoptic, historically Hegic/Premia), TVL is LP capital directly at risk. On a vault (Rysk, Stryke-style structured products), TVL is deposits running a defined strategy.

We will not publish a single TVL-based ranking across all of these models — it would compare numbers that don’t mean the same thing. See our Onchain Options Rankings for how we actually order the venues we’ve reviewed.

The venue map (2026)

• Derive — CLOB options and perpetuals, formerly Lyra, the venue most public research credits with the large majority of onchain options premium. V3 (proposed, not live) would move execution to a zkVM with Ethereum settlement. Real concentration risk for the whole category sits here.
• Paradex — CLOB plus Paradigm’s institutional RFQ network landing onchain; the credible story for existing professional block flow moving onchain instead of growing retail from zero.
• Aevo — hybrid options, perps, and pre-launch markets on its own L2, descended from the Ribbon Finance DOV lineage; real product, smaller options-specific share than Derive by most public accounts.
• Rysk — options repackaged as income: deposit into a covered-call or cash-secured-put style vault instead of managing a book. Live on Arbitrum and HyperEVM. You are the writer by default.
• Panoptic — an options-like payoff built directly from Uniswap concentrated-liquidity positions, with a design claim of no external price oracle. A genuinely different mechanism, not a ranked competitor to a CLOB.

A handful of other venues are worth knowing even though they don’t fit the ranking above — each gets its own full review: Stryke routes orders into CLAMM, orderbook, and RFQ paths (including a stated Paradigm RFQ integration) rather than running a book of its own. Convallax prices calls and puts on event probabilities instead of an asset’s spot price, sitting deliberately on the border between options and prediction markets. CallPut is a new venue tied to the GIWA (Upbit-linked L2) narrative, too early to know its real open interest. Opyn / Squeeth is the original power-perpetual (ETH²) design — the historical reference for what a genuinely expiry-free “perpetual option” actually is. Premia offers American-style options via concentrated-liquidity AMM pools, real and live but a sliver of volume next to Derive. Hegic was one of the earliest onchain options protocols, a peer-to-pool AMM now mostly of historical interest.

A few more names keep coming up but aren’t reviewed yet, because there isn’t enough to confirm confidently: Hypercall (a pivot toward 0DTE-style options, thinly documented as of this writing), SOFA (referenced alongside the structured-vault layer), scattered 0DTE-style tokenized-equity options projects on Base, and copycat “DRV flywheel” tokens on HyperEVM that have no actual connection to Derive — worth knowing about as a confusion/scam risk, not as somewhere to put money.

Not on this list at all: Hyperliquid, which is a perp venue (see Perp DEX Rankings) — an options product there (widely discussed as “HIP-5” on CT) is unconfirmed as of this writing and lives on our watchlist instead. Deribit is the centralized benchmark we compare against, not a row in an onchain table.

Tokens and farming — keep this separate from the product

Several venues here have an associated token or points program, and CT discussion around the category leans heavily token-first. Treat the product and the token as separate questions: does the venue actually work the way it claims (matching, settlement, risk handling), and separately, does the token capture value from that in a way you can verify — not “the token is pumping, therefore the product must be good.” The same discipline we apply to pre-TGE perp DEX farming applies here: farm only what you’d genuinely use anyway, and don’t treat an undated points program as a guaranteed future airdrop.

Risks, front and center

As a buyer: you can lose 100% of the premium you paid. That’s the option working as designed, not a malfunction — time decay will erode a long position that doesn’t move in your favor in time.

As a writer, LP, or vault depositor: you’re being paid to warehouse volatility. A gap move, an oracle failure, or a thin book during stress can produce losses well beyond a “yield” headline number would suggest. Peer-to-pool and vault designs concentrate this risk directly onto LPs and depositors.

Venue risk: smart-contract bugs, sequencer or (for Derive’s proposed V3) zkVM issues, incorrect settlement, admin-key risk, and incentive cliffs when a rebate or points program ends — all the standard onchain-protocol risks, plus options-specific settlement complexity on top.

Market-structure risk: with one venue credited with the large majority of onchain options premium, a Derive-specific outage or a contested governance decision is a category-wide event, not just a venue-specific one.

Oracle and collateral risk: options on LSTs, wrapped assets, pre-launch tokens, or (eventually) RWAs inherit whatever depeg or oracle risk those underlying assets carry, on top of the options mechanism itself.

What to actually watch through Q4 2026

We keep a running version of this on our Options Watchlist, but the short version: whether Derive’s V3 migration ships cleanly, whether Paradex/Paradigm RFQ prints persist as a trend rather than a launch spike, whether Hyperliquid ships a real options market, whether tokenized-asset (RWA) options listings become real size anywhere, and whether the category’s share of Deribit-scale notional actually moves — or stays roughly where public comparisons put it in 2026.

Where to start if you’ve never traded an onchain option

Don’t start by writing. Paper the payoff of a single weekly call or put on a small size before risking real capital either direction. If what you actually want is yield, read a vault’s documented max-loss scenario before its advertised return — the return number is the easy part to market, the loss scenario is the part that matters. If you want active flow and are choosing a venue, Derive (screen-based CLOB) and Paradex (RFQ/structures) are, per this guide’s research, the two most established venues as of this writing — verify current fees, liquidity, and audit status directly on each before committing capital.