The restaking sector has entered its value-accrual phase. After 18 months of TVL competition, the three leading restaking protocols — EigenLayer (now rebranded EigenCloud), Symbiotic, and Karak (now rebranded OpenGDP) — have taken radically divergent paths on the question that matters most to inv...
The restaking wars aren't about TVL anymore. They're about which corporate entity captures protocol revenue — and whether token holders ever see a dime of it.
The restaking sector has entered its value-accrual phase. After 18 months of TVL competition, the three leading restaking protocols — EigenLayer (now rebranded EigenCloud), Symbiotic, and Karak (now rebranded OpenGDP) — have taken radically divergent paths on the question that matters most to investors: where does the money go?
EigenLayer's ELIP-12 proposal, published in December 2025, represents the most consequential governance shift in restaking history. It introduces a 20% fee on AVS rewards and channels 100% of EigenCloud service revenue into a buyback contract for EIGEN holders. If implemented, this transforms EIGEN from a governance-only token into a fee-accruing asset — the restaking equivalent of Uniswap's historic fee switch.
Meanwhile, Symbiotic — backed by Paradigm and Lido co-founders — remains deliberately tokenless and governance-minimized, operating with non-upgradeable contracts and $1.44B in TVL. And in the most dramatic pivot, Karak has abandoned the restaking narrative altogether, rebranding to OpenGDP and repositioning as a Layer 1 for programmable GDP and stablecoin payments. The XP staking program is winding down, and the $KAR token's purpose is in limbo.
For institutional allocators, the implications are clear: EigenLayer is making the strongest bid for token-holder value accrual, but it comes with significant dilution risk (36.8M EIGEN unlocked February 1, more coming March 1). Symbiotic offers the purest restaking exposure but no token to capture that value. And Karak/OpenGDP is a completely different thesis than what investors originally signed up for. The niche players — Tangle Network, Catalysis, and Puffer Finance — are where the real alpha may be hiding.
Development activity tells a more honest story than marketing. Here's what the code reveals:
EigenLayer (Layr-Labs/eigenlayer-contracts): 712 stars, 468 forks, 84 open issues. The most recent commits (February 2026) focus on CI/CD improvements and Claude AI-powered code review integration — signaling maturity and operational tooling rather than net-new feature development. The core slashing infrastructure shipped in April 2025 and the contract base is now feature-complete. Activity has shifted from core protocol development to ecosystem tooling: the eigensdk-rs Rust SDK saw commits as recently as February 9, 2026, indicating investment in developer experience for AVS builders.
Symbiotic (symbioticfi/): Active across multiple repositories. The relay-contracts repo (11 stars, 14 forks) pushed a release on February 13, 2026 — the day of this report. The brand-new rewards-v2 repo saw its README updated the same day, suggesting an imminent rewards system overhaul. Critically, Symbiotic's GitHub activity is concentrated in its official symbioticfi org, while the community ecosystem is thin — no equivalent of EigenLayer's awesome-avs curation layer.
Karak: Virtually no public GitHub activity. The karak-network/v2-contracts repo exists but shows minimal recent updates. This aligns with the OpenGDP rebrand — development resources appear to have shifted away from restaking contracts toward the new L1 infrastructure.
Emerging Signal — Tangle Network: The tangle-network/blueprint repo (333 stars, actively maintained) pushed a "TOML-based per-job pricing config" feature on February 12, 2026. This is significant: Tangle is building a pricing engine that lets operators set granular fees per job type — a level of economic sophistication that EigenLayer's operator-AVS model currently lacks.
EigenWatch Subgraphs: A new repo (EigenWatch/subgraphs) committed on February 13, 2026, building real-time monitoring for slashing events, delegation metrics, and AVS operations. This kind of infrastructure middleware signals that EigenLayer's ecosystem is maturing beyond core protocol development into analytics and risk management tooling — a prerequisite for institutional adoption.
The restaking sector's total addressable TVL stands at approximately $12.7 billion across the three major protocols:
| Protocol | TVL (Feb 2026) | Market Share | Token Status | |----------|---------------|--------------|--------------| | EigenLayer/EigenCloud | ~$11.1B | ~87% | EIGEN (live, declining) | | Symbiotic | ~$1.44B | ~11% | No token | | Karak/OpenGDP | ~$260M | ~2% | KAR (announced, not launched) |
EigenLayer's dominance is overwhelming but misleading. TVL peaked near $19.5B in late 2025 and has contracted roughly 43% — likely driven by a combination of EIGEN's 91% token price decline, the expiration of airdrop-farming incentives, and rational capital reallocation as competing yield opportunities emerged. The protocol still commands 87% market share, but the trend is directionally negative.
Symbiotic's $1.44B TVL is more impressive than it looks. The protocol has been at deposit capacity for extended periods, suggesting unmet demand that artificial caps deliberately constrain — a Lido-style approach to managing growth responsibly. The collateral base is genuinely diversified: wstETH, ENA (Ethena), and sUSDe compose the largest deposits, demonstrating multi-asset restaking in practice, not just in whitepapers.
Karak's $260M TVL represents a dramatic decline from its peak and reflects the strategic uncertainty around the OpenGDP rebrand. Existing restakers face an awkward question: did they deposit assets to secure distributed services, or to underpin a stablecoin payments L1?
ELIP-12, published by the Eigen Foundation in December 2025, is the most important governance proposal in restaking history. It introduces three mechanisms that fundamentally change EIGEN's value proposition:
1. AVS Rewards Fee (20%): On stake subsidized by EIGEN incentives, the newly formed Incentives Committee will collect a 20% fee on AVS rewards. This fee flows into a smart contract designed for buybacks — creating direct deflationary pressure on EIGEN supply as AVS adoption grows.
2. EigenCloud Revenue Capture (100%): All revenue from EigenDA, EigenCompute, and EigenAI — after operator expenses — flows to the same buyback contract. This is the critical innovation: it ties EIGEN's value directly to actual cloud service revenue, not just speculative staking demand.
3. Productive Stake Prioritization: Only fee-paying AVSs qualify for staker and ecosystem incentives going forward. This forces a separation between "productive stake" (actively running live services with slashing exposure) and "passive stake" (parked capital farming airdrops). The implications for capital allocation are profound.
The corporate reality behind EigenLayer is a familiar Web3 trifecta:
The ELIP-12 fee model creates a genuine — if imperfect — value loop for token holders. However, the Incentives Committee is staffed by the Foundation and Labs, not elected by token holders. This is governance theater with real economic consequences: the entities that control fee allocation are the same ones that raised $164M in equity.
EIGEN faces persistent dilution. On February 1, 2026, 36.8 million EIGEN tokens (~$12.3M) unlocked, representing over 2% of total supply and 6.75% of market cap. The next unlock hits March 1, 2026. Approximately 591.9 million tokens have been unlocked to date, with an infinite total supply model creating perpetual dilution. For a token that has already declined 91% from peak, this supply overhang is existential.
EigenLayer's slashing mechanism went live on mainnet on April 17, 2025, completing the protocol's original vision. The Unique Stake Allocation system allows operators to earmark portions of delegated stake for specific AVSs, ensuring only one service can slash that particular stake. Launch partners include LayerZero (for cross-chain messaging enforcement) and Infura's Decentralized Infrastructure Network. The 14-day withdrawal waiting period creates a built-in slashing window, but also introduces capital inefficiency that liquid restaking protocols like Ether.fi ($3.8B TVL) and Renzo ($2.9B TVL) partially address.
Symbiotic represents the opposite end of the governance spectrum from EigenLayer. Backed by a $5.8M seed round led by Paradigm and Cyber Fund (the Lido co-founders' investment vehicle), Symbiotic employs non-upgradeable core contracts that deliberately eliminate governance attack surfaces.
The founding team — Misha and Algys from Statemind, a top-tier DeFi auditor with deep Lido ties — designed the protocol around five core primitives: collateral, vaults, operators, resolvers, and networks. Each component is modular and composable, giving network builders unprecedented control over their restaking configuration without requiring protocol-level governance votes.
Here's the uncomfortable truth: nowhere, yet. Symbiotic has no live token. The Points Program explicitly states that points hold "no intrinsic cash or monetary value" and can be revoked at any time. The protocol generates no direct revenue for external stakeholders.
Value accrues to:
The Lido-Paradigm nexus behind Symbiotic is a corporate governance story disguised as protocol governance. The same investors who back Lido (the dominant liquid staking provider) also back its restaking partner. This vertical integration creates a powerful flywheel for stETH deposits but concentrates economic control in a small number of entities.
Symbiotic has secured partnerships with LayerZero, Ethena, Marlin Protocol, Frax, Redstone, and Ether.fi. The first partner cohort is focused on oracles, cross-chain messaging, and sequencer infrastructure — use cases where Symbiotic's modular architecture provides genuine differentiation over EigenLayer's more opinionated design.
In what may be the most dramatic strategic pivot in DeFi history, Karak has rebranded to OpenGDP and repositioned as "the global base layer for programmable GDP." The team describes it as "the evolution of Karak. Same team, same codebase lineage, clearer mission."
The new focus areas — tokenization, exchange, and coordination — bear little resemblance to the original restaking thesis that attracted $48M in funding from Coinbase Ventures and Pantera Capital. The OpenGDP Network Stack (ONS) enables custom blockchain networks with shared security, but the emphasis has shifted from securing crypto-native AVSs to underwriting real-world economic activity like stablecoin payments and asset tokenization.
For prospective KAR token holders, the pivot introduces fundamental uncertainty:
This is a case study in how corporate-level strategic decisions can strand token holders who invested in one thesis and woke up in another. The Coinbase and Pantera-backed entity controls the roadmap; XP point holders have no governance recourse.
Tangle Network (333 GitHub stars, daily commits) is building what may be the most sophisticated economic model in restaking. Its Blueprint system — analogous to EigenLayer's AVSs — includes a built-in fee distribution model: 50% to developers, 30% to operators and restakers, 20% to the protocol treasury.
The February 12, 2026 commit introducing "TOML-based per-job pricing config" means operators can set granular fees per job type — a level of pricing sophistication that EigenLayer's flat-rate model cannot match. For institutional operators running multiple service types, this granularity is a prerequisite.
Tangle's focus on complex offchain computations — MPC, zero-knowledge proofs, AI inference — positions it as the restaking layer for compute-intensive workloads, complementary rather than competitive with EigenLayer's broader ambitions.
Catalysis has pivoted to institutional DeFi coverage, leveraging restaked capital from EigenLayer, Symbiotic, and SatLayer to provide over $20B in protection capacity. This is a 100x increase over existing coverage protocols and addresses a critical gap: only 0.5% of DeFi TVL is currently insured.
The model is elegant: idle restaked ETH, BTC, and SOL are delegated to specialized CoverPools managed by professional curators. Institutions apply for coverage, pay fees through smart contracts, and receive automated payouts. This transforms restaked capital from a passive yield play into active insurance underwriting — generating real revenue from real risk pricing.
For token governance analysts, Catalysis represents the "middleware thesis" for restaking: the value isn't in the restaking protocol itself, but in the specialized applications built on top of restaked security.
Puffer Finance ($1.4B TVL at peak) has developed one of the more interesting governance models in the liquid restaking space through its vlPUFFER system. Users stake PUFFER tokens to mint vePUFFER NFTs, with voting power accruing over time without requiring fixed lockup durations. Unstaking resets accrued voting power, incentivizing long-term participation while maintaining flexibility.
The H2 2026 roadmap includes an AI-driven Puffer AppChain Marketplace and MCP models for the UniFi AppChain — an ambitious expansion from liquid restaking into application-specific infrastructure. Recent GitHub activity (January 2026) shows LayerZero bridging integration and pufETH deployment on megaETH, indicating multi-chain expansion.
| Protocol | Revenue Source | Who Captures Value | Token Holder Share | Grade | |----------|---------------|-------------------|-------------------|-------| | EigenCloud (EIGEN) | AVS fees (20%), Cloud revenue (100%) | Buyback contract → EIGEN holders | Direct (via buybacks) | B+ | | Symbiotic | None currently | Paradigm, Cyber Fund (equity) | None (no token) | N/A | | Karak/OpenGDP | Undefined post-pivot | Coinbase Ventures, Pantera (equity) | Unknown | D | | Tangle | Blueprint fees (30% to restakers) | Operators, developers, treasury | Direct (30% share) | A- | | Catalysis | Insurance premiums | CoverPool curators, restakers | Indirect (via restaking yields) | B | | Puffer | Restaking spread + AppChain fees | vlPUFFER governance participants | Indirect (governance power) | B- |
The winner on pure value accrual is Tangle Network, with its transparent 30% fee share to operators and restakers. EigenLayer's ELIP-12 is a close second, but the buyback mechanism introduces execution risk (the Incentives Committee controls timing and magnitude) and the 91% token decline suggests the market hasn't priced in the fee switch yet — or doesn't trust it.
ELIP-12 is the most important governance proposal in restaking: The 20% AVS fee + 100% cloud revenue buyback mechanism gives EIGEN its first real value-accrual story. Watch the Protocol Council ratification vote closely.
Symbiotic's governance-minimized design is a feature, not a bug — but it means no token value accrual. The Paradigm-Lido vertical integration creates significant corporate value, but external token holders have no vehicle to participate.
Karak's rebrand to OpenGDP is effectively a rug on the restaking thesis. Existing XP holders should treat their points as worthless until the team provides concrete token utility under the new mandate.
Tangle Network offers the most transparent fee-sharing model in restaking. The 50/30/20 split (developers/restakers/treasury) and per-job pricing engine are institutional-grade economic design.
Catalysis shows restaked capital's killer app may be insurance, not additional yield. The $20B coverage capacity unlocks institutional capital that has been sitting on the sidelines.
The liquid restaking layer (Ether.fi, Renzo, Puffer, Kelp) is where most users actually interact with restaking. Ether.fi at $3.8B and Renzo at $2.9B have more user-facing significance than the base layer protocols.
Token unlock pressure on EIGEN is severe. With 36.8M tokens unlocking February 1 and more on March 1 against an infinite supply model, the buyback mechanism from ELIP-12 would need to generate substantial revenue to offset dilution.
Cascading slashing risk: If a vulnerability affects multiple AVSs secured by the same restaked ETH pool, total slashed collateral could trigger systemic contagion. The September 2025 Ethereum slashing event (39 validators) demonstrated that correlated operator failures are a real threat, per CoinDesk.
Governance concentration in ELIP-12: The Incentives Committee is staffed by Eigen Foundation and Eigen Labs — the same entities that raised $164M in equity. Token holder governance is downstream of corporate decision-making.
Symbiotic token risk: If and when Symbiotic launches a token, early Paradigm/Cyber Fund allocations could create massive insider selling pressure. The points program's explicit disclaimer ("no intrinsic value") provides no legal protection.
Regulatory exposure: The SEC's evolving framework on staking services could reclassify restaking protocols as unregistered securities offerings, particularly where fee-sharing mechanisms (like ELIP-12's buybacks) create investment contract dynamics.
Client concentration risk for EigenDA: Mantle Network's migration from EigenDA to Ethereum blobs demonstrates that AVS clients can leave, taking revenue with them. EigenCloud's revenue projections depend on sticky demand that hasn't been proven.
Karak/OpenGDP execution risk: A pivot from restaking to programmable GDP requires rebuilding market positioning, developer tooling, and user trust from scratch. The $48M in funding may not be sufficient.
The restaking wars are no longer about who has the most TVL. They're about who builds the tightest loop between protocol revenue and token holder value.
EigenLayer wins this round. ELIP-12's fee switch — 20% on AVS rewards, 100% of cloud revenue into buybacks — is the strongest value accrual mechanism in the restaking sector. But the corporate structure dilutes this story: the Incentives Committee is appointed, not elected; Eigen Labs' $164M equity raise means shareholders may ultimately capture more value than token holders; and the infinite token supply model creates perpetual dilution that buybacks must overcome.
Symbiotic is the best protocol with the worst token story (because there is no token). Its governance-minimized, Paradigm-backed architecture is technically superior for network builders, but the value flows exclusively to equity holders and the Lido ecosystem.
Karak/OpenGDP has left the chat. The restaking thesis is dead here; what emerges from the OpenGDP rebrand is a different investment entirely.
The real alpha is in the niche players: Tangle Network's transparent fee-sharing model, Catalysis's insurance-as-a-service for restaked capital, and Puffer's flexible vlPUFFER governance represent the kind of structural innovations that mainstream coverage misses. The $3,000/month client who reads beyond the blue chips will find their edge here.
The restaking sector is maturing from a TVL arms race into a revenue war. The protocols that survive will be the ones that answer one question clearly: who gets paid?