Bunni

  • Name: Bunni
  • URL: https://bunni.pro/
  • Tags: ethereum-ecosystem
  • Category: liquidity-incentive infrastructure / shapeshifting concentrated-liquidity AMM / vote-escrowed DEX governance layer
  • Summary: Bunni is best understood as a grafting of Curve-style gauge politics onto Uniswap-style concentrated liquidity, later extended into a Uniswap-v4-native AMM with automated liquidity-shape control. In v1, Bunni wrapped Uniswap positions into fungible ERC-20 LP tokens so gauge voting, ve-token boosts, and option-style liquidity incentives could sit on top of concentrated liquidity. In v2, Bunni pushed the control surface further upward with Liquidity Density Functions, autonomous rebalancing, rehypothecation, an auction-managed AMM, and a zone contract that can restrict rebalancing execution to approved solvers. The reusable insight is that Bunni turns LP incentives and liquidity-shape management into explicit governance and infrastructure layers rather than leaving them as passive consequences of an underlying AMM.
  • What it does:
    • Wraps Uniswap liquidity positions into fungible ERC-20 tokens so multiple LPs can share the same position and use it in token-native DeFi plumbing
    • Uses Curve-inspired gauges and vote-escrowed governance so lockers can direct incentives toward specific Bunni liquidity pools
    • Uses oLIT option tokens instead of plain token emissions, turning liquidity mining into a discounted continuous token-sale mechanism that routes more cash back to the protocol
    • In Bunni v2, lets LPs choose or deploy liquidity shapes through Liquidity Density Functions instead of static concentrated-liquidity ranges alone
    • Adds autonomous rebalancing, rehypothecation of idle assets, and an auction-managed AMM intended to recapture MEV and reduce LP leakage
    • Splits core authority across immutable pool contracts, governance-selected zone infrastructure, and solver pathways used for automated rebalancing
  • Key claims:
    • The official docs describe Bunni as a liquidity engine for incentivizing Uniswap liquidity and say its wrapper turns Uniswap positions into ERC-20 tokens, which is what makes Curve-style gauges possible on top of otherwise NFT-like LP positions
    • The v1 docs explicitly say Bunni adopted Curve’s gauge system for Uniswap positions and that gauges distribute oLIT based on veLIT voting
    • The docs say Bunni uses a Balancer 80LIT-20WETH LP token as the lock asset for veLIT, raises the max reward boost above Curve’s default, and uses option-style reward tokens instead of direct emissions
    • The gauges docs say Bunni can kill out-of-range gauges using a dedicated oracle, so incentive routing is conditioned on the protocol’s own view of whether a position remains in range
    • The v2 docs describe Liquidity Density Functions, shapeshifting liquidity, autonomous rebalancing, rehypothecation, and an am-AMM as first-class protocol features rather than peripheral strategy add-ons
    • The v2 technical overview says BunniZone restricts auto-rebalancing execution to a whitelisted solver set plus the am-AMM manager, and that governance can choose a new zone version for all pools; this is a meaningful control surface above the immutable pool logic
    • The public bunni-v2 repository warns that Bunni v2 was exploited and that the repository does not include the relevant fixes, which is an important caution when treating the codebase as a live canonical implementation
    • Bunni’s later exploit postmortem shows that its custom active-vs-idle liquidity accounting and Liquidity Density Function estimation were not just implementation details but core solvency surfaces; see bunni-v2-liquidity-density-function-idle-balance-rounding-liquidity-rebound-sandwich-drain-2025-09-02
  • Whitepaper: ../whitepapers/bunni-v2-whitepaper.pdf and ../whitepapers/bunni-primary-sources-2026-05-08.md
  • Sources:

Internal linkages

Control surface

  • The leverage sits in veLIT vote concentration, gauge eligibility, oracle choices about whether gauges remain live, zone-version governance, and solver whitelists for rebalancing.

  • The pool substrate matters, but user outcomes can still be steered materially by the higher-layer control plane around it.

  • Last reviewed: 2026-06-27 UTC