Sailfish vs Midgard
This page addresses common questions from those discovering Sailfish for the first time, including how it compares to Midgard - another Layer 2 solution for Cardano with very different goals.
Disclaimer
This comparison reflects our understanding to the best of our knowledge, based on the current Midgard specification and publicly available progress. Midgard is under active development and details may change; we'll revise this page as new information becomes available.
At a glance
Sailfish is a modular Layer 2 optimized for DeFi - specifically, intent-based smart account DeFi protocols. It removes friction for traders and liquidity providers by making order placement, cancellation, and pool management free signed messages, with settlement to L1 handled in batches via an Echo-validated snapshot.
Midgard is a general-purpose Layer 2 designed to let existing Cardano dApps scale with minimal code changes. It uses an optimistic rollup model with bonded operators and fraud proofs, prioritizing compatibility and a familiar developer experience.
Each design comes with real tradeoffs. Sailfish trades general-purpose dApp compatibility for a streamlined DeFi experience. Midgard aims to be a more holistic solution, but is more difficult to bootstrap and suffers from liquidity fragmentation.
Development status
| Sailfish | Midgard | |
|---|---|---|
| Mainnet readiness | Mainnet ready | Under active development; not yet deployed to a testnet. Current ETA: end of 2026 |
Comparison
| Sailfish | Midgard | |
|---|---|---|
| Purpose | Scaling DeFi features | General-purpose scaling |
| Smart Account support | Yes | No |
| Supports intents | Yes | No |
| New token required | No | No |
| Transaction cost | Gasless Intent | L2 Transaction Fee |
| Settlement finality on L1 | Cardano L1 finality after next snapshot - ~12 min (~36 blocks, high-probability) or ~12 hr (k=2160, full) | 3–7 days (maturity_duration) - optimistic fraud-proof challenge window |
| Emergency self-exit | 24-hour on-chain timelock - no operator or coordinator needed | No direct user exit; recovery requires escape_hatch to attract new operators at reduced bond |
| Front-running / MEV | Structurally impossible - settlement batches have no per-transaction ordering to exploit | Standard rollup MEV - the shift's operator has full ordering control |
| Who orders transactions | Deterministic rules - proposers can't cheat; the committee refuses invalid proposals | The shift's bonded operator (1-hour shifts), kept honest by fraud-proof challengers |
| Trust model | Echo committee threshold-signs every settlement; the on-chain validator only accepts signed settlements. Safety depends on a majority of honest Echo operators, but is still protected as long as a single honest coordinator is present | Permissionless bonded operators (50K–200K ADA each); anyone can submit a fraud proof to slash a bad block. Safety depends on sufficient honest bonded operators or active fraud-proof challengers |
| Coordination | Modular - Sailfish handles execution/settlement, Echo verifies snapshots, a single coordinator currently proposes them. Decentralizing the coordinator is a planned upgrade | Holistic - directory, scheduler, block production, settlement, and recovery bundled into one protocol |
| Liveness | Gated by the snapshot coordinator - settlement halts if it goes offline. User funds are always accessible via manual withdraw | Depends on the active bonded operator producing blocks; halts until escape_hatch brings in new operators |
| Liquidity fragmentation | None - Sailfish intents, AMM pools, and L1 DEX liquidity all reachable via a single matcher route | Multi-level - L2 isolated from L1 (multi-day withdrawal to bridge), and within L2 depends on each dApp |
| Liquidity routing | Unified - auto-aggregated across Sailfish intents, AMM pools, and L1 DEXes | Per-dApp; no cross-L2 routing |
| Funds locked while in DeFi | No - funds stay in your Smart Account; any funds in use rely on a soft reservation mechanism | Depends on each dApp, usually yes |
Bottom line
Sailfish is a DeFi-optimized L2 built around intents and Smart Accounts. It removes most of the friction associated with on-chain trading - no per-action fees, no fragmented liquidity, no multi-day withdrawal windows - at the cost of being narrower in scope. Sailfish optimizes for safety of funds over liveness, using the same trust model as Echo: the on-chain emergency exit guarantees user funds remain self-custodial regardless of network state.
Midgard is a general-purpose L2 that prioritizes compatibility, giving existing Cardano dApps a path to scale without rewriting their code. The tradeoff is a per-action fee model, multi-day withdrawal windows tied to the optimistic challenge period, and liquidity that is fragmented.
Sailfish is live on Mainnet, while Midgard is under active development with the current ETA being end of 2026.