Radiant Capital launched in 2022 as a cross-chain lending protocol combining Aave v3 smart contracts with LayerZero’s interoperability layer, aiming to let users deposit and borrow assets seamlessly across Arbitrum, BNB Chain, and Ethereum. In October 2024, a severe exploit (later attributed to North Korea-linked attackers) drained roughly $50M, and the protocol never recovered. On June 1 2026 the Radiant DAO voted to wind down operations: active development and RDNT emissions stopped, and borrowing was disabled across all markets. This review examines Radiant’s architecture, security posture, and its wind-down.
What it is
Radiant Capital is a cross-chain money market protocol that lets users supply and borrow assets across multiple EVM chains without bridging. It clones Aave v3’s core logic and uses LayerZero’s messaging for cross-chain state synchronization. The project originally positioned itself as an “omnichain” lending hub, with RDNT token incentives driving liquidity. Following the October 2024 exploit, the protocol paused most operations, and after failing to recover the stolen funds, secure new investment, or sustain its treasury, the DAO voted on June 1 2026 to wind down. Borrowing is disabled, RDNT emissions have ended, and only immutable contracts and a victim-recovery portal remain.
How it works
Radiant’s architecture mirrors Aave v3, meaning users deposit assets into lending pools and receive interest-bearing rTokens. Borrowers can take out loans against their deposits, subject to overcollateralization ratios. Uniquely, Radiant uses LayerZero to coordinate pool state, deposit caps, and liquidation parameters across chains, so a user on Arbitrum could, in theory, borrow from liquidity sitting on BNB Chain. In practice, the protocol originally supported cross-chain borrowing and repayments via LayerZero’s generic message passing. After the June 2026 wind-down, borrowing is disabled and only immutable contracts remain; the protocol is no longer actively operated. The RDNT token functions as a governance and incentive token, with early emissions heavily subsidizing liquidity providers. All operations are governed by the Radiant DAO through token-weighted votes on upgrade proposals and parameter changes.
Key numbers
TVL stands at approximately $1M (DeFiLlama, as of 2026-07-15), down more than 99% from a peak of roughly $386M before the hack. Radiant was deployed on three chains: Arbitrum, BNB, and Ethereum. Three audit firms — PeckShield, Zokyo, and BlockSec — reviewed the protocol’s contracts prior to launch. The project launched in 2022. The October 2024 exploit drained roughly $50M, and on June 1 2026 the DAO voted to wind down operations.
Security and audits
Before deploying, Radiant engaged PeckShield, Zokyo, and BlockSec for code audits, covering both the Aave v3 fork and the LayerZero integration. These audits failed to prevent the October 2024 exploit, which exploited a vulnerability in the cross-chain messaging logic. Details remain sparse, but the exploit resulted in the immediate shutdown of most markets and a drain of user funds. No bug bounty program is publicly known for Radiant, and its smart contracts are currently frozen except for the limited safe-mode pools. Governance is managed by the DAO; no off-chain multisig override is publicly documented. The lack of a transparent post-mortem and the incomplete remediation plan heighten user uncertainty. As far as public records show, there have been no subsequent incidents since the safe-mode activation.
Strengths
1. Cross-chain foundation: The integration of Aave v3 with LayerZero’s cross-chain messaging remains technically sound in principle. Radiant was among the first to deliver truly omnichain lending, with deposits and borrows spanning Arbitrum, BNB, and Ethereum.
2. Multiple audits: Three independent firms (PeckShield, Zokyo, BlockSec) reviewed the codebase, which at least signals an initial commitment to security, even though the exploit occurred.
3. Orderly wind-down: Rather than abandoning users, the DAO kept immutable contracts and a victim-recovery portal live, so any funds recovered from the October 2024 attackers can still be returned to affected users.
Weaknesses and risks
1. Major exploit and shutdown: The October 2024 exploit drained roughly $50M and shattered confidence. TVL collapsed from a ~$386M peak to under $1M, and on June 1 2026 the DAO voted to wind the protocol down entirely.
2. Wind-down: Active development and RDNT emissions have stopped and borrowing is disabled across all markets; the protocol is effectively decommissioned, retaining only immutable contracts and a recovery portal.
3. No recovery: After failing to recover stolen funds, secure new investment, or sustain its treasury, the DAO chose to wind down rather than rebuild. Affected users may never be made whole, and the protocol will not return as a going concern.
How it compares
Radiant’s TVL of under $1M places it far behind major lending protocols. Aave Aave holds roughly $14B TVL with a multi-year security track record and formal verification from Certora. Morpho Blue Morpho Blue (~$7B) offers permissionless isolated markets with custom oracles, avoiding the systemic risk that plagued Radiant. Compound Compound (~$1.3B) maintains simpler but robust single-asset borrowing markets. Radiant’s cross-chain ambition was genuinely innovative, but after the exploit, that complexity proved fatal. Euler V2 Euler V2 ($1B) demonstrates a successful recovery from a similar oracle exploit, rebuilding to a comparable TVL with modular vaults, while Radiant chose to wind down rather than rebuild. Although Radiant underwent three audits (PeckShield, Zokyo, BlockSec), Aave's continuous code reviews and formal verification provide greater assurance. With the DAO having voted to wind down rather than reopen markets, Radiant cannot be compared favorably to any peer.
Verdict
Radiant Capital was a bold cross-chain experiment that collapsed under the weight of its own complexity. The 2024 exploit and its lingering consequences—a decimated TVL, a failed recovery, and a June 1 2026 DAO vote to wind down—leave the protocol effectively decommissioned. While the underlying idea and initial audit efforts merit recognition, the current reality is that Radiant offers neither the security nor the utility of its competitors. DeFi Intel rates Radiant Capital 4.5 out of 10.