The Complete Guide to Onchain Options

Onchain options are still a small and concentrated market. Public research currently points to Derive as the largest onchain venue by premium volume, while total activity remains far below the main centralized benchmark, Deribit. Treat that concentration as a category risk, not just a leaderboard detail.

This guide covers the parts that matter before you trade: how options work, how venues differ, what the main numbers mean, and where buyers and writers can lose money. Start with the payoff, then compare execution, collateral, settlement, and current liquidity.

Options in one page — no Greeks required

An option gives the buyer the right, but not the obligation, to buy or sell an asset at a set strike price. A call is the right to buy. A put is the right to sell. You pay a premium for that right.

The premium is the contract price. It is not the strike price, and it is not a guaranteed profit. For example, paying a $40 premium for a call means the underlying must move far enough above the strike to cover that $40 and any other costs.

	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

Options can look like a midpoint between perps and prediction markets. Like a prediction contract, an option has a defined payoff. Like other markets, its price changes with time, volatility, and the underlying asset. That is why options take longer to learn than a simple long-or-short perp.

The payout comes from the other side of the trade. That may be an individual writer, an LP pool, or a structured product that took on the risk. A venue matches and settles the trade; it does not create yield from nowhere. An income product pays a seller a premium for accepting a defined obligation.

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

Earlier protocols such as Opyn, Hegic, Ribbon-style DeFi Option Vaults, and the first AMM-based Lyra faced four problems. L1 gas made weekly positions expensive to manage. Better-informed traders could exploit weak AMM pricing. Thin markets made it hard to find writers. Perps also offered a simpler way to trade direction.

By 2026, newer venues were using cheaper chains, order books, and RFQ systems. Those designs can make multi-leg trading and quoting easier than a pure AMM. The category is still much smaller than perps and centralized options, so treat growth claims as a live thesis rather than a settled fact.

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 or structured product, TVL is deposits running a defined strategy; that comparison does not describe Rysk’s current independent RFQ positions.

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 — RFQ-based covered calls and cash-secured puts: choose the asset, target price, expiry, and collateral, then receive an upfront premium if you accept a quote. Its current product documentation describes independent, fully collateralized positions rather than a pooled vault; verify live deployments and exit rules.
• 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.

Other venues matter because they use different designs, not because they belong in one ranking. Stryke routes orders through several liquidity paths. Convallax sits closer to event markets than standard asset options. CallPut is early and needs deployment and liquidity checks. Opyn / Squeeth is a historical reference for expiry-free designs. Premia and Hegic use pool-based models with different liquidity and writer risks.

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 structured-product seller: 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, vault, and RFQ-based income designs expose the seller to the payoff and settlement risks documented by that product.

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.