Skip to content

Benchmarks

Side-by-side against the closest comparables. Numbers come from competitor docs and on-chain reads. Where Atrium loses, we say so.

DimensionAtriumCascade (Solana)August (Solana)
ChainArbitrum + RH-Chain (when SDK)SolanaSolana
Margin computeStylus (Rust, ~10-100x cheaper compute-heavy ops, per Arbitrum Stylus)Native Solana programsN/A
Open adapter standardIPorticoAdapter v1.0 (MIT)ClosedClosed
Agent integrationSigil EIP-712 + Postern session keys + ERC-8004LimitedNone
Formal verification5 Kani+proptest invariants in CIAudit onlyAudit only
UX polish todayFoundation: Verifier Mode complete, app surfaces in progressMore matureMore mature
Mainnet liveNo, testnet build phaseYesYes

Honest read: the Solana-native comparables have UX polish and mainnet liveness Atrium does not yet match. Atrium leads on chain choice (EVM is larger), open standards, agent-layer depth, and formal verification. The polish gap closes as the app surfaces complete; the standards gap holds.