The opinions expressed are my own and not influenced by Outlier Ventures. This investment thesis does not constitute investment advice. Parts of this were written with the help of AI language models to improve clarity and save time. The thesis is written using publicly available, non-material information.

Written: 18. Sep | Last edited: 13. Nov | Published: 20. Nov

Special thanks to Marco Cora from the ZKsync Foundation for taking the time to answer my questions and share additional information, which helped improve the accuracy of this thesis.

Executive Summary

ZKsync is evolving from a single L2 into an Elastic Network of ZK chains that share proving, routing, and a Gateway on Ethereum. This positions ZKsync as core middleware for privacy, compliance, and low-latency interoperability, with native account abstraction improving UX. The narrative is shifting from raw throughput to privacy, compliance, and verifiability. ZKsync already ships account abstraction, a production zkEVM, the ZK Stack for custom chains, and fast proving with Airbender. Early ecosystem signals, Deutsche Bank’s compliance L2 via the ZK Stack, and Prividium for banks, suggest real product-market fit where public L1s struggle with cost and data exposure.

• ◆Atlas upgrade: Introduces a high-performance sequencer (>15 k TPS, up to ~40 k TPS in tests), the Airbender RISC-V zkVM with ~1-second proofs and minute-scale L1 finality, and full bytecode-level EVM equivalence through a multi-VM compiler. Together, these make Ethereum the real-time liquidity hub for all ZK Stack chains, enabling sub-second L2↔L2 interop and institutional-grade settlement.
• ◆Proven fee base: ~$60m collected in early operation before EIP-4844 cut L1 DA costs, cost leadership from state-diff efficiency, margins should improve as blob pricing normalizes with rising rollup usage.
• ◆Native interop tolls: Gateway and Shared Bridge create a per-message fee on cross-chain routing, a $0.001 toll on a legacy-rails proxy of 1.6-1.9 trillion messages per year implies a ~$1.6-1.9 billion annual revenue opportunity, before growth.
• ◆Prover and DA economics: ongoing prover cost declines and higher throughput shorten L1 finality and expand fee margins, with EigenDA-class partners lowering data-availability costs.
• ◆Enterprise ARR: Prividium-style deployments monetize via per-chain subscriptions plus usage-based interop and DA or proving, a single bank running 100-200 chains implies ~$10-20 million ARR per bank before message tolls.
• ◆ZKnomics linkage and token logic: Under the “From Governance to Utility” framework, ZK wants to move from a pure governance token to an active network utility asset. When governance activates the fee switches and public burn, revenues from the Gateway, Shared Bridge, and sequencer flow to the Token Assembly, which buys ZK on the market and allocates it between burns, staking rewards, and ecosystem growth, with splits set by governance.
• ◆Institutional RWA rails: ZKsync enables compliant issuance with selective disclosure, auditor/regulator visibility, permissioned venues, and paymaster-sponsored gas, with ~$3.5B “assets secured” cited, including ~$1.7B from Tradable private credit; Prividium provides bank-grade blockchain-as-a-service so institutions can run one or many private ZK chains anchored to Ethereum; Deutsche Bank, under Project DAMA 2, is building a compliance L2 for tokenization, payments, and fund operations. Procurement is SLA-backed, churn is low, and deployments expand across jurisdictions and business lines. Success should be measured by issuance throughput and recurring sequencing, proving, and DA fees rather than TVL.
• ◆Account abstraction and conversion moat: protocol-level AA and paymasters let issuers sponsor gas for KYC’d users, raising conversion and adding non-speculative fee flow.
• ◆Ecosystem traction: More teams are launching chains on the ZKsync stack, with recent announcements from ADI Chain, SandChain, and Xsolla.

• ◆Elastic interop with fast finality: one wallet and same address across rollup, validium, and volition, minute-scale settlement so cross-chain feels single-domain.
  

What to watch

• ◆Decentralized delivery: sequencing and proving not yet decentralized.
• ◆Interop usage and latency: cross-chain messages per day, median time to L1 finality
• ◆Confirm activation: Governance-approved utility mechanics (fee switches, burn, staking permissions) and resulting token buyback or burn activity.
• ◆Unit economics: prover cost per transaction, DA cost per transaction
• ◆Enterprise adoption: permissioned chains live, issuance throughput, chains per institution
• ◆Ecosystem traction: live ZK Stack chains, monthly active wallets, share of AA or sponsored gas.
• ◆Product traction: Gravity volumes and institutional share, share of account-abstracted or sponsored transactions
• ◆Competitive pressure: OP Stack, Polygon CDK, Arbitrum, Scroll, Linea, Starknet.
  

Positioning and horizon: Accumulate on a long-term view, on the premise that zk-powered privacy and interop go mainstream. Entry timing is uncertain, so build gradually and size for a multi-year hold. Reassess if fee switches and burn are materially delayed, if decentralized sequencing stalls, or if institutional pilots fail to convert into live issuance, payments, and fund operations.

The model is already validated, with ZKsync generating around $60M in revenue so far. Revenue is still relatively modest today, but that is part of the appeal: if the economics were fully proven and scaled, much of the upside would likely be priced in. The technology stack is in place, and the team of 90+ people, roughly 70–75% of whom are engineers, continues to push it forward. The next phase is about signing more chain partners, standing up those chains, and letting them scale, which should compound into a network effect as interop volume grows and additional revenue streams accrue from new chain deployments.

Positioned Beneath the Stack

ZKsync offers a multi-product ZK infrastructure that sits beneath the user layer, which makes its value proposition difficult to convey to a general audience. It’s challenging to articulate why it’s superior or how its technology will shape the future. However, as it matures and its ecosystem develops, the story is gradually becoming clearer. Over time, I believe this growing clarity will allow the market to better grasp ZKsync’s potential and drive broader adoption.

The project sits in the infrastructure stack, beneath the user-facing layer of wallets, dApps, and consumer interfaces, building core middleware infrastructure to power an elastic network. Most users don’t yet interact with or even see this layer, but it powers much of what will happen above it. As ZKsync moves closer to the consumer layer and showcases real-world applications, the market will start recognizing why ZK and privacy are so critical to blockchain usability.

We’re already seeing early signs of this narrative shift. As the industry begins to appreciate the role of privacy and the need for scalable, verifiable systems that aren’t fully transparent by default, ZKsync’s approach is likely to gain significant mindshare. In short, its story is still unfolding, but as the stack matures and awareness of ZK’s importance grows, I expect ZKsync to have its moment.

Modular Architecture

ZKsync’s design follows a modular systems approach instead of a single monolithic chain built to serve every use case. It forms an Elastic Network of sovereign chains that share proofs, interoperability, and liquidity through a common Gateway on Ethereum. Each component, including the sequencer, prover, and data-availability layer, can evolve independently while remaining cryptographically unified.

This architecture enables independent optimization, fault isolation, and specialization across layers. Each ZK Stack chain can choose its own sequencer set, fee token, and data-availability provider while inheriting Ethereum’s security. The main trade-off, cross-chain interoperability, is addressed through validity-proved messaging and same-address UX, allowing applications across different ZKsync chains to compose as if they operated on a single network.

Execution environments such as public rollups, institutional Prividiums, or appchains scale horizontally while retaining shared verification and liquidity. Users interact through a unified wallet and address space, realizing the modular ideal of invisible interoperability.

ZKsync reconciles two long-standing design goals: the composability of monolithic chains and the scalability of modular stacks. It gives developers the sovereignty of appchains without sacrificing distribution or liquidity. As modular data-availability and interoperability layers mature, ZKsync’s Ethereum-anchored model becomes the coordination fabric connecting specialized execution domains under a shared framework of verifiable security and real-time liquidity.

How will value accrue back to the ZKsync Network?

On-chain fundamentals

For Era, and any ZK Stack chain that adopts ZK as the base staking token:

• ◆Fee base already proven: with roughly $60m in fees generated in the first months of operation before EIP-4844 blobs reduced L1 DA costs by ~99 percent.
• ◆Relative cost leadership: historically the cheapest rollup due to state-diff efficiency.
• ◆Reversion to advantage: as blob pricing normalizes with rising rollup usage, ZKsync’s data-efficiency edge should reassert itself where optimistic rollups cannot match
• ◆Validity-proof compression: improving margins on sequencing and proving.

Interop economics

ZKsync’s Gateway and Elastic Network create a per-message toll model for cross-chain and cross-domain routing. As a volume proxy, global financial rails such as SWIFT, SEPA, FedWire, and DTCC process an estimated 1.6–1.9 trillion messages per year, growing roughly 15 percent CAGR over the past decade. As a TAM illustration, a $0.001 per-message toll implies $1.6 - 1.9 billion in annualized interop revenue potential before growth. If even a subset of messages can be charged as a percentage of value flowing rather than a flat toll, this number could easily be at least an order of magnitude higher. Legacy message volumes have historically grown very fast, and we can reasonably expect crypto-native messages to grow even faster, driven by always-on global liquidity, demand for instant settlement, and the shift from batch payments to real-time, programmable transactions.

Why growth should outpace legacy rails

• ◆Network flattening, firms, and even individuals progressively act as their own banks, increasing machine-to-machine and agent-to-agent messaging.
• ◆Instant payments demand, settlement expectations compress toward real-time, lifting message frequency.

Why chain count will explode

• ◆App-sovereignty: successful dApps seek to control their lane, customize throughput, fees, and MEV policy, which we already see with Hyper, Uniswap, and others.
• ◆Enterprise and regulatory partitioning: no single chain fits every jurisdiction or business line. A large bank may need separate chains per country and per division (IB, Wealth, AM, Commercial), quickly reaching hundreds of instances.

UX moat: value accrues only if the network feels like one chain: one wallet, same address, no network switching, and the ability to call contracts on any chain from any chain. This requires native ZK interop to preserve security and latency without third-party bridges.

As blob markets equilibrate and interop volume scales, sequencing fees, DA/proving payments, and per-message routing tolls become durable revenues. With ZKnomics, these flows are designed to route to staking and burn, tying usage directly to the ZK token.

Off-chain Fundamentals

Enterprise willingness to pay: Banks and large enterprises already budget for SLA-backed infrastructure, compliance tooling, and support. Prividium-style deployments monetize through an annual subscription model per private chain plus usage-based fees for interop, DA, and proving. Banks and other institutions would need several chains, each of its legal entity can run its own chain per jurisdiction, aligned with local regulations, since it is effectively impossible to build one global chain that satisfies every country’s rule set. A group like Deutsche Bank could end up operating hundreds of chains, with cross-entity coordination handled via ZK messages. While there is currently a single, non-sharded gateway, as the system scales, the operator can deploy multiple gateways and shard them, making the network more resilient to downtime and failures.

Revenue lines

• ◆Per-chain subscription: managed sequencing and proving, EigenDA quota, SLA and support, compliance modules.
• ◆Interop tolls: Gateway, and Shared Bridge message fees, billed in stablecoins.
• ◆Paymasters and gas sponsorship: issuer-funded transactions for KYC’d users.
• ◆Professional services: integrations, migrations, audits, and training.
  

Buyback and allocation under ZKnomics: As proposed in ZKnomics, protocol revenues can be routed to the Token Assembly, which can programmatically acquire ZK and then allocate to burns, staker rewards, or growth proposals, aligning cash flows with token value and network security.

Illustrative scale: If a single Prividium instance is priced at ~$100k per year, and a universal bank operates 100-200 chains across jurisdictions and divisions, that implies $10-20 million ARR per bank before any per-message interop tolls. This is an example, not a forecast, but it shows how multi-instance footprints can compound.

Why this is durable: Regulated buyers value policy enforcement, auditability, and privacy, procurement cycles are slower, but churn is low, expansions follow a land-and-expand pattern across business lines and countries.

Key dependencies: Turn on fee switches and burn in governance, maintain cost leadership in proving and DA, keep UX unified so the Elastic Network feels like one chain.

Many see a potential conflict of interest between Matter Labs and the token, but revenue flows to the DAO, which then decides how much to pay Matter Labs, and Matter Labs can, in principle, be replaced.

Graphic

Tokenomics

The ZK token launched as the native governance asset for the ZK network. Token allocation is structured as follows: 33% allocated to employees and investors, of which ~5–6% has already been distributed, and ~0.8% unlocks monthly to investors such as Parafi, CoinFund, Variant, Lightspeed, and Faction. Roughly 6 B tokens will vest over four years to backers, including Placeholder, a16z, Dragonfly, Blockchain Capital, and USV. Separately, 17.5% was allocated to the airdrop, 19.9% to the foundation, and 29.6% to the Token Assembly, a governance body which oversees major unlocks or expenditures (holders receive a 30-day exit window before execution).

Graphic

The shift from governance to utility: ZK is now proposed to evolve into a governance + utility asset under a framework dubbed ZKnomics. According to the “From Governance to Utility: ZK Token Proposal, Part I” document, as the network matures, with interoperability, institutional rails (Prividium), and multiple chains live, the token model must capture real value. The key foundations of the redesign are: (1) making decentralization economically sustainable; (2) capturing a meaningful share of the value the network creates; and (3) aligning all incentives around the token.

Value capture vectors:

• ◆On-chain interoperability fees: When assets or messages move across ZK Stack chains or between public and private domains, modest protocol fees apply. These fees represent the first credible native capture mechanism.
• ◆Off-chain enterprise licensing: Advanced modules for compliance, reporting, treasury integrations, or audit interfaces used by regulated institutions are monetized, and some revenue is routed back to the network economy.
  

Token-economic mechanism: All value captured via the above vectors flows into a governance-controlled engine. Proceeds are used to:

• ◆Buy ZK tokens from the market (token buybacks)
• ◆Allocate across: (i) token burns to reduce supply, (ii) staking rewards for decentralized operators, and (iii) ecosystem treasury for protocol and ecosystem growth
• ◆This mechanism creates a self-reinforcing loop: network usage → value capture → token buyback/burn → strengthened network/incentives.
  

Implementation roadmap: The rollout is explicitly staged and contingent on the delivery of key infrastructure and governance approval. Major milestones include:

• ◆Authorizing permissionless staking (via ZIP)
• ◆Upgrading the token contract with a public burn mechanism
• ◆Enabling sequencer and interoperability fee switches
• ◆Finalizing allocation rules via governance vote
  Until these stages are complete, many utility mechanics remain dormant and would activate only when usage and infrastructure maturity meet expectations.
  

Risks & watch-points:

• ◆Execution risk: The token model depends on uptake of interoperability features and enterprise modules; if volumes remain low, value capture could lag.
• ◆Governance: Parameter setting (fee levels, buyback size, burn splits) remains in community hands and may evolve unpredictably.
• ◆Decentralization trade-offs: As value flows back into token incentives, maintaining “credible neutrality” is critical and non-trivial.
• ◆Market timing: The economics assume scale (inter-chain traffic, institutional adoption), which may take time to materialize.
  

With the revamped tokenomics, ZK is shifting from a voting tool to the economic engine of an interoperable, institution-ready multi-chain network. The proposal gives the token a clear purpose: capture value from usage and enterprise activity, route that value into buybacks/burns and rewards, and thus reinforce decentralisation and network strength. All of these mechanisms, staking activation, burn enablement, fee routing, and allocation rules, remain subject to governance approval by token holders, with stakers (or their delegates) directly voting through the Token Assembly on each stage before activation.

Ecosystem

On its website, ZKsync reports 18 chains on its network, more than 700 million transactions processed, securing over $4 billion in total value, with a throughput capacity of over 15,000 transactions per second (~0.75 Ggas/s) per chain, and maintaining 99.99% uptime.

Ecosystem Structure

ZKsync is transitioning into an “Elastic Network”, a cluster of specialized ZK chains that share core infrastructure, including common proving and routing, while ultimately settling to Ethereum.

An Elastic Network represents a web of ZK-powered chains that operate as one unified system. Each chain can independently choose between rollup, validium, volition, or Prividium data availability, allowing projects to tailor performance, privacy, and cost. Assets and messages move natively through Ethereum-based infrastructure, ensuring no reliance on third-party bridges and delivering a consistent UX across all chains.

By vertical, the network serves: 

• ◆ZKsync Era: Consumer and infrastructure hub
• ◆Abstract, Sophon, Lens, SandChain, XSolla: Entertainment, social, and consumer applications
• ◆Memento (Project Dama): Deutsche Bank’s institutional chain
• ◆Gravity: Trading and financial apps, optimized for speed and privacy
• ◆ADI Chain: sovereign and real-world asset infrastructure layer
  

Developers can launch app-specific chains using the ZK Stack, customizing execution logic, fee models, data availability, and governance. Shared proving and native interoperability ensure a seamless user experience across the network.

Operational costs typically fall in the tens of thousands of USD per month, depending on configuration, for example, prover access, throughput, and data availability requirements.

For users, one wallet works across the entire Elastic Network, with unified signing and account abstraction creating a single, consistent interface across Abstract, Gravity, Lens, Sophon, Memento, and future ZK Stack deployments.

Recent chain announcements.
Activity on the ZKsync stack has not stalled, with Xsolla launching a ZKsync-based chain for game-native payments and asset infrastructure, SandChain aligning its launch with the Atlas upgrade to improve throughput and builder UX, and ADI Chain live in testnet backed by major UAE institutions for a dirham-pegged stablecoin and sovereign-grade tokenization.

Looking Ahead: High-Conviction Verticals Within the ZKsync Ecosystem

I believe there are many more applications and verticals that can drive value to the ZKsync ecosystem, such as Abstract, but for the sake of brevity, I’ll focus on a few areas I’m excited about.

ADI Chain - Sovereign Stablecoin & Infrastructure Layer

ADI Chain is a sovereign-grade blockchain network on the ZKsync Stack, developed by the ADI Foundation (IHC’s digital arm) in partnership with ADQ, First Abu Dhabi Bank, and IHC, chaired by Sheikh Tahnoon bin Zayed. It introduces a dirham-pegged stablecoin issued by FAB as the financial backbone for cross-border payments, digital identity, and real-world infrastructure tokenization across emerging markets, starting in Africa.

The chain leverages ZKsync’s zero-knowledge rollup architecture for scalability, privacy, and regulatory compliance, while integrating EigenDA for high-throughput data availability. Its mission is to bring 1 billion users into the digital economy by 2030 through interoperable rails connecting sovereign capital, banks, and enterprises to Web3.

Strategic role in the ZKsync network:

• ◆Acts as the sovereign and RWA infrastructure layer, linking public ZKsync rails to national payment systems and regulated institutions.
• ◆Expands ZKsync adoption beyond DeFi, embedding stablecoin-based payments, identity verification, and energy/commodity settlement into real-world economic flows.
• ◆Complements Gravity (finance), Memento (institutional), and Era (consumer) as the emerging-market and sovereign counterpart in the Elastic Network.

Gravity - Perp and Spot Dex

Gravity is a privacy-first onchain finance platform on the ZKsync Stack, centered on a ZK spot and perp DEX plus an investment marketplace. It raised $19M Series A on Sep 18, 2025, co-led by ZKsync, Further Ventures, EigenCloud, and 500 Global. 

Under the hood, it runs a ZKsync Validium L2, privacy by default, with L2 state validated without publishing raw data, Ethereum verifies transactions on L1 via ZK proofs, scalability and costs improve through batching and proof posting, and EigenDA supplies high-throughput data availability at roughly 100 MB/s with programmable privacy on the roadmap.

The core problem in on-chain perps is privacy, open positions are visible, predators hunt liquidations, and institutions with size retreat to CEXs. Gravity solves this with faster, cheaper, private trading, ~200-300 ms confirmations, on-chain dark-pool style execution for institutions, and ZK-verified off-chain matching that avoids on-chain order-book inefficiencies. Net, it is positioned as a central liquidity hub in ZKsync’s Elastic Network.

How Gravity compares to Lighter, on privacy and its trade-offs: Gravity is privacy-first by design, running as a ZKsync Validium appchain that keeps trade, position, and liquidation data off-chain while settling state transitions to Ethereum with ZK proofs, and it’s integrating external DA like EigenDA to scale, which protects institutional flow from adversarial scraping but makes withdrawals depend on the off-chain DA path if the operator fails. Lighter takes the opposite stance, prioritizing transparent, provably fair execution, it’s a ZK rollup that publishes compressed state deltas to Ethereum blobs so users can reconstruct state and, if the sequencer misbehaves, trigger a contract-level Exit Hatch for self-withdrawal, meaning weaker privacy than Gravity, stronger permissionless exit guarantees. 

AI Agents on ZKsync

As AI systems become more autonomous, trust and verifiability move from nice to have to mandatory. ZKsync’s zero-knowledge architecture lays the groundwork for verifiable AI agents, where model usage, dataset integrity, and policy compliance can be proven on chain without exposing sensitive data.

Using ZK proofs, agents can attest to which model they ran, which dataset they drew from, and how an inference was executed. The result is an auditable trail that preserves privacy, yet confirms correctness. In collaboration with Eigencloud, ZKsync is advancing proof-of-correctness for AI, often framed as “HTTPS for AI”: heavy computation happens offchain, then compact proofs anchor trust on-chain.

This unlocks three practical benefits:

• ◆Compliance by construction, agents must satisfy verifiable constraints before acting, from model version and dataset provenance to risk and KYC policies.
• ◆Safety and integrity, proofs reduce manipulation and hallucination impact by forcing actions to pass explicit checks.
• ◆Selective visibility, regulators, and partners can verify outcomes while prompts, weights, and raw data remain private.

By integrating verifiable computation into the ZKsync Elastic Network, AI agents move from “trust me” to “verify me.” As these capabilities surface in user-facing applications, expect adoption to accelerate, because the system becomes private where it matters, and connected where it counts.

Real World Assets On-Chain

Not all tokenized RWAs should be openly tradable on public rails; by design, many must be permissioned, gated, or even kept internal to support compliance, operational control, and later transferability. ZKsync reportedly verifies about $3.5B of such assets, but the headline TVL is less important than why these assets sit on-chain and how the rails monetize. 

Of the roughly $3.5B “assets secured” figure often cited, Tradable accounts for about $1.7B across nearly 30 institutional private credit positions issued on ZKsync. In this model, assets are secured and tokenized on ZKsync, while access is controlled at the platform layer for KYC, AML, and accredited-investor requirements. 

Success should be measured by compliant issuance throughput, fee economics paid to ZKsync for sequencing, proving, and data availability, and the depth of permissioned market activity, rather than aggregate TVL alone.

I believe ZK provides a strong foundational technical layer and clear PMF for institutions that want to tokenize assets. BlackRock and others have shown issuance works fine on Ethereum L1, but recurring mints, burns, transfers, and daily NAV updates face high, volatile gas, more MEV exposure, and no native account abstraction for smooth, KYC’d UX. A ZK L2 like ZKsync provides Ethereum-grade security with lower, predictable costs, fast finality, and built-in account abstraction so issuers can sponsor gas in stablecoins. 

Privacy is better via selective disclosure with zk-KYC, proof-of-eligibility, and permissioned markets enforcing allowlists, while the ZK Stack enables private, policy-controlled chains that still interoperate with Era. ZK offers a strong foundation and real PMF for institutional tokenization, pairing Ethereum security with low-cost settlement, a compliance-friendly stack supporting allowlists, paymasters, and permissioned venues, plus Chainlink-powered NAV and pricing feeds that integrate into custody, treasury, and reporting. 

With the ZK Stack and Gateway enabling default interop, tokens distribute across wallets, neobanks, OTC venues, and DeFi for conservative collateralization and structured products, lowering operating costs and expanding the buyer base, with key risks being regulation, oracle reliance, and ZKsync’s path to fuller decentralization.

Deutsche Bank is collaborating with ZKsync technology to develop a compliance-focused Layer 2 blockchain solution for asset tokenization, as part of its Project Dama 2. This initiative, developed with partners including Memento Blockchain and Interop Labs, uses ZKsync's ZK stack to create a private and scalable environment that meets financial regulations while improving transaction efficiency. This ZKsync-powered L2 chain is designed to provide features like curated validators, regulatory oversight, and KYC-gated access to facilitate the tokenization of funds and other assets within the traditional finance system.  

Prividium

Prividium is a ZK-based blockchain platform secured by Ethereum, purpose-built for regulated institutions that need privacy, compliance, and operator control. It delivers private execution with public verifiability; transactions remain confidential while succinct proofs of correctness are anchored to Ethereum. Through ZKsync’s Elastic Network, enterprises deploy private, compliant chains that interconnect with Ethereum and the broader ZK ecosystem without bridges, and auditors and regulators get selective, role-based visibility.

Prividium originated as a banking product for Deutsche Bank and then evolved into a general platform. Deutsche is launching a compliance-focused Layer 2 using the ZKsync stack as part of Project DAMA, built with partners such as Memento Blockchain and Interop Labs. The design prioritizes AML, KYC, and suitability today; tokenization and fund lifecycle automation follow next. ZK proofs certify that policies were enforced without exposing raw data, so institutions get practical compliance, not just theoretical cryptography. 

Each institution controls its own chain, its KYC perimeter, and who can see its internal state. Inside the perimeter, markets can be permissionless for admitted participants; outside parties only learn what is selectively disclosed. Curated validators and role-based permissions govern deployment rights, transaction ordering, and data access. Users are issued compliance artifacts, for example, an exobound NFT that encodes identity checks and suitability, and products are classified so clients can only interact with instruments appropriate to their profile. Every batch generates ZK proofs, Ethereum finalizes the state, and no sensitive data leaves the operator’s environment.

Prividium unlocks

• ◆Cross-border payments between regulated entities, settlement in minutes, compliance enforced on chain, data stays private.
• ◆Digital asset issuance and lifecycle, issuance, distributions, redemptions, and audit trails anchored to Ethereum; internal details stay confidential.
• ◆KYC-anchored marketplaces, peer-to-peer trading among verified participants, lower counterparty and data-leak risk.
• ◆Secure cross-chain finance, atomic swaps, and institutional workflows that move value between chains with proofs, not intermediaries.
  Intraday liquidity and repo, real-time collateral and cash movements with policy gates and auditable controls.

Prividium combines Ethereum-grade security, enterprise-grade privacy, and enforceable compliance in production. Banks get private where it matters, connected where it counts, and a clear path from today’s policy-first deployments to tomorrow’s tokenized markets.

Technology

ZKsync is a Layer-2 scaling solution for Ethereum that uses zkRollup technology to achieve high throughput and low fees without compromising on security. As a validity rollup, every off-chain batch is accompanied by a zero-knowledge proof that Ethereum verifies, ensuring the batch is correct. This removes the need to trust operators, the Ethereum base layer checks the proof, making the system trustless and secure by design.

ZKsync was among the first to achieve full EVM compatibility in a ZK rollup. Developers can deploy Solidity or Vyper smart contracts on ZKsync Era without rewrites, enabled by a custom LLVM-based compiler and zkEVM. In practice, ZKsync Era is a general-purpose, EVM-compatible zkRollup, which lowers migration friction for dApps and combines Ethereum’s developer experience with the performance benefits of zk proofs.

How ZKsync’s zkRollup works: User transactions are bundled into off-chain batches. A zkSNARK or zkSTARK prover generates a concise proof that all transactions obey Ethereum’s rules. The validity proof is posted to Ethereum, where a ZKsync contract verifies it and finalizes the batch. Because invalid batches cannot finalize, unlike optimistic rollups that rely on later fraud proofs, ZKsync offers fast finality, withdrawals can be completed in minutes once the proof is accepted on L1.

Core components and products

ZKsync Era (zkEVM rollup): A general-purpose zkRollup on Ethereum with smart contracts, near-instant confirmations, and low fees. It uses validity proofs for security and achieves Ethereum-level finality after L1 verification. Era maintains interoperability with standard Ethereum tooling. zkPorter, a planned hybrid validium data path, can further boost scalability, the current focus remains on the rollup.

Developer SDKs: Multi-language SDKs, for example, TypeScript, Python, Swift, Rust, abstract interaction with the network. They cover wallets, transaction signing, contract calls, bridging, and features like account abstraction and L1-L2 messaging, reducing integration effort and lowering entry barriers.

ZKsync External Node: A read-replica node anyone can run. It fetches blocks, replays, and verifies them locally from genesis, allowing trustless reconstruction of Era’s state. The node exposes standard Web3 APIs for queries and transaction submission. It does not propose blocks, the sequencer is currently centralized, but it verifies the sequencer’s work, improving transparency and enabling independent auditing.

ZK Stack and custom chains: An open-source framework to launch new zk chains, L2 or L3, interoperable with ZKsync Era. The core Era code is open-sourced under permissive licenses with templates for sovereign chains that reuse the zkEVM and proving system while allowing customizations, for example, sequencer sets, permissioning, and fee tokens. Native interoperability is built in, chains connect into an “Elastic Network” via a shared L1 hub and standardized proofs, passing messages and assets trustlessly without third-party bridges. The result is horizontal scaling, multiple chains that share liquidity and communicate in minutes, with Era as the first chain and others, including private instances, joining the rollup cluster anchored to Ethereum.

Figure: Conceptual architecture of the Elastic Network, multiple ZKsync chains, rollups, and validiums, run in parallel as a rollup cluster with a shared L1 bridge for validation and data availability. Common proofs and messaging enable trustless interoperability while each chain can be customized or kept private, all under Ethereum security.

Era smart contracts on Ethereum: A set of L1 contracts manages rollup state commitments, verifies validity proofs, and handles bridging. The core rollup contract receives batches and new state roots, verifies proofs, and updates the official L1 record. Auxiliary contracts, for example, Bridge Hub or Message Hub, coordinate cross-chain messaging and asset movement across ZK Stack chains and Ethereum. A shared bridge provides finality and interoperability for all instances, enabling trustless inter-chain transfers and anchoring off-chain activity to Ethereum.

Technical advantages

Airbender prover, speed, and throughput: Airbender is ZKsync’s latest zkSNARK prover, engineered for very fast, low-cost proofs. Benchmarks show up to 21.8 MHz, proving throughput on a single NVIDIA H100, over 6× faster than several alternative zk proving systems. In practice, Airbender can prove a full Era block in about 35 seconds on one GPU, pushing proving costs toward roughly $0.0001 per simple transfer and enabling sub-second L2 confirmations. Faster proving shortens time to L1 finality, improves UX, and points toward near real-time settlement and future “home-made” proofs on consumer hardware. The result is high throughput with low fees while preserving decentralization of proving.

Near real-time interoperability: Fast proofs plus a unified bridging system let ZKsync chains communicate and transfer assets with low latency. Hyperchains connected by hyperbridges use shared proofs and L1 contracts, so cross-chain messages feel like single L2 transactions, minutes rather than days, no third-party validators or 7-day waits. This enables an internet of verifiable, interconnected chains and avoids the latency and risk profile of external bridges.

Built-in privacy and compliance, Prividium: Institutions can run private zkRollups connected to Ethereum. Only commitments, state hashes, and proofs are posted on L1, underlying transaction data stays within the institution. This yields practical privacy, sensitive details remain private while proofs certify correctness. Protocol-level compliance features include role-based access, KYC or AML integration, and audit logs. Selective disclosure or viewing keys can give regulators controlled visibility. Each institution sees only its own chain, confidentiality is preserved, yet chains can interoperate via ZKsync Connect when desired. This combines privacy, auditability, and interoperability, a middle ground between public transparency and permissioned silos, and supports enterprise pilots such as bank deployments.

Ethereum-level security and decentralization roadmap: As a rollup, ZKsync inherits Ethereum security, L1 validators verify proofs, so compromising ZKsync would require breaking the cryptography or Ethereum itself. The codebase is open-source, and the project is working toward decentralizing sequencing and proposing. External nodes and community provers already add transparency. With advances like recursive SNARKs and Ethereum data blobs for cheaper data availability, ZKsync is positioned to evolve into a fully decentralized, community-operated network, aligned with the view that zk rollups are the long-term path for Ethereum scaling.

Competition and Comparables: where ZKsync sits among ZK L2s and privacy rails

ZK L2 peers

Starknet: ZK rollup using zk-STARKs (no trusted setup, quantum-secure but larger proofs). Runs the Cairo VM with native account abstraction. Not natively EVM; Solidity needs transpilers like Warp. Strong prover roadmap (recursive STARKs, STWO prover), sustained high-double-digit TPS, batch finality in tens of minutes. No protocol-level privacy or KYC. Sequencing and proving decentralizing; fees in STRK. Read-through vs ZKsync: better UX primitives and STARK tech depth, but non-EVM friction and STRK gas reduce adoption and institutional comfort.

Scroll: Type-2 zkEVM (bytecode-equivalent to Ethereum) using Halo2 and hardware acceleration; DA on Ethereum with blobs. Full EVM equivalence, same tooling. Seconds-level sequencing, proofs in tens of minutes. No native privacy. Decentralized prover/sequencer roadmap; gas in ETH. Read-through vs ZKsync: both now offer “it’s just Ethereum” bytecode-level compatibility, but Scroll leans into minimal-diff migration, while ZKsync differentiates with interop, native AA, and a stronger privacy layer.

Linea (ConsenSys): zkEVM (SNARK) with full EVM parity, blob DA, and 5k TPS roadmap. Deep MetaMask/Infura integration. Minutes-scale L1 finality, hundreds of thousands of daily tx. Public protocol, enterprise pilots via partners (notably SWIFT’s interbank trial). Consortium-style governance; gas in ETH, dual ETH + LINEA fee burn. Read-through vs ZKsync: Linea has a stronger enterprise brand and infra distribution; ZKsync is stronger in native privacy and permissioned ZK-chain design while now matching EVM bytecode equivalence.

Polygon zkEVM: Type-2 zkEVM (PLONK). On-chain DA now, validium/Avail in future, integrated into Polygon 2.0. High EVM fidelity, ETH gas. Proofs every ~20 min, low fees but low usage. Public, no built-in privacy; enterprise work via other Polygon products. Stage-0 governance, Polygon-run operators, POL staking to come. Read-through vs ZKsync: Polygon gains scale from a large ecosystem, but ZKsync already executes the ZK-first, privacy + interop thesis Polygon is still converging toward.

Optimistic L2 baseline

Arbitrum: Optimistic rollup with multi-round fraud proofs, single sequencer, 7-day challenge window. Perfect EVM equivalence and the largest L2 liquidity base. Instant UX confirmations, high throughput, but delayed trust-minimized withdrawals. Public chain with DAO governance (Stage 1 decentralization). Read-through vs ZKsync: liquidity and reliability benchmark, but ZKsync’s validity proofs give faster exits and a more natural path to verifiable compute and privacy.

Privacy-first institutional rails

Keeta (institutional L1): Purpose-built DAG + DPoS ledger with Google-class infra, targeting database-like speed and compliance-first design. Not EVM, no general-purpose smart contracts; exposes a rules engine and tokenization primitives via SDKs. Multi-million TPS with ~400 ms confirmation, trading decentralization for performance. Protocol-level KYC and selective disclosure; permissioned validators and consortium governance. Read-through vs ZKsync: Keeta wins for banks needing on-ledger policy enforcement and sub-second finality; ZKsync’s Ethereum-anchored Prividium offers similar compliance guarantees while retaining composability with public liquidity.

How ZKsync stacks up

With the Atlas upgrade and subsequent v27 EVM-equivalence release, ZKsync now delivers ~1-second proofs via the Airbender RISC-V zkVM, >15k TPS throughput, and full bytecode-level EVM equivalence through an EVM interpreter running on EraVM, so Solidity contracts can be deployed without custom tooling.The ZK Stack connects chains with sub-second interop and direct access to Ethereum liquidity, effectively turning Ethereum into the real-time capital hub for its network. Combined with protocol-level privacy (Prividium) and a modular path to decentralized sequencers and provers, ZKsync, in my opinion, is positioned as one of the most institution-ready ZK L2s, bridging public and private domains under Ethereum security.

ZKsync combines private execution with public verification, same-address interop, and a credible decentralization roadmap into the most coherent end-to-end ZK stack. With full EVM equivalence and native privacy rather than bolt-on confidentiality, it is one of the most institution-ready public L2s. The market’s next phase is clearly tilting toward privacy, compliance, and verifiable interop, and ZKsync is architected exactly for that, pairing a privacy-first, interop-native design with early RWA and institutional pilots. If the team executes on decentralization and long-term economics, ZKsync can evolve into the neutral middleware layer that both private and public chains route through for secure, compliant, real-time settlement.