Polygon Labs’ proposal for Celo to become an ETH L2 using Polygon Chain Development Kit ( CDK )

Polygon Labs is excited to welcome Celo home to Ethereum in the chain’s transition from an independent EVM-compatible L1 to an Ethereum L2. As Ethereum devotees obsessed with scaling, we are beyond excited to see Celo expand into an emerging and thriving Ethereum ecosystem and help extend Ethereum blockspace.

We would like to offer an alternative path for transitioning to an L2, one that would let Celo maximize the advantages of being an Ethereum L2 while maintaining the characteristics that made the chain a success in the first place. Migrating to an Ethereum L2 is a large technical upgrade that should come with built-in Ethereum alignment and as-close-as-possible developer experience, without compromising security.

To that end, Polygon Labs proposes that Celo Ecosystem choose to deploy an L2 using Polygon CDK (Chain Development Kit), an open source toolset developers can use to spin up their own ZK-powered L2 for Ethereum–and that permits existing L1s to convert to Ethereum L2s with little friction and an emphasis on modularity.

Here is an overview of the key value propositions for deploying using Polygon CDK:

  • Increasing cross-community collaboration by joining hands in an ecosystem of robust, ZK-powered L2s;
  • An EVM-equivalent environment for enhanced compatibility, so developers can experience the closest alignment with Ethereum as technically possible;
  • Increased security that leverages Ethereum’s battle-tested consensus layer with the security of zero-knowledge proofs, instead of relying on social-economic incentives;
  • Maintain low fees, made possible by a zkEVM validium architecture and off-chain data availability;
  • Access to Unified L2 Ecosystem that brings together Ethereum mainnet and Polygon ecosystem into one massive economy, with the feel of a single chain;
  • Customizability of an app-chain stack and interoperability with all Polygon L2 chains;
  • Near-instant withdrawals, compared to a week;
  • Fast finality and instant cross-chain interactivity, secured by ZK; and
  • Support that meets the standards of the biggest Web3 projects and Web2 companies alike.

This document will cover Polygon Labs’s alignment with Ethereum and development of its ZK tech stacks, deep-dive into the value propositions for upgrading to ZK, suggest a possible design, and walk through the impact on key stakeholders.

Our proposal is the beginning of a dialogue in a scenario where, no matter what, everyone wins, because Celo will be transitioning to an Ethereum L2!

Ethereum Alignment and Background in ZK

Ethereum is the paragon of decentralized, permissionless technology, and every step taken by Polygon Labs has been with the goal of scaling not only Ethereum, but its values, as well. With L2 and rollups on its roadmap, Ethereum’s future relies on implementing the best technology to scale securely and in a frictionless way.

Core developers at Polygon Labs have spent years building ZK technology that aligns with Ethereum as closely as possible. Breakthroughs, including the Plonky2 ZK proving system, have meant that L2s see reduced fees and finality times across use cases, all while inheriting the decentralization, security, and existing tooling and smart contracts that make Ethereum the best programmable blockchain.

To fully align with Ethereum’s vision of scaling with on-demand L2 deployment, we developed Polygon CDK (Chain Development Kit). Chains deployed using Polygon CDK extend Ethereum blockspace. These chains join a unified L2 ecosystem with the fastest canonical L2 <> L2 transfers. In the future, these chains will experience automatic shared liquidity through a novel ZK interoperability layer. That means that all L2s developed with CDK are connected to all others, creating unified liquidity in an environment that feels like a continuous mesh of blockspace.

In practice, Polygon CDK means near-instant cross-chain interactivity, Ethereum included, all with the power of ZK proofs. This is an unlock made possible specifically with ZK, and unavailable to fraud proofs. Chain states are aggregated in a single layer and posted to Ethereum, the validity of which is guaranteed by math, rather than social-economic incentives.

The result is a network of ZK L2s that are so seamlessly connected it feels like a single chain, with one-click access to Ethereum’s liquidity.

Chains being developed using Polygon CDK include Immutable zkEVM, Gnosis Pay, Canto, Capx, Palm Network, and others, as well as Polygon PoS (with a proposal to become a zkEVM validium), Polygon zkEVM, and (eventually) Polygon Miden.

Polygon Labs believes in an equitable, inclusive world brought forward by decentralized technology, and our belief is that such a world can be achieved with alignment within a broader Ethereum community.

Upgrade Value Propositions

The following is broken into two parts:

  1. Addressing the initial value propositions proposed by cLabs; and
  2. Describing unique value propositions unlocked by Polygon CDK.

At a high-level, these responses walk through the technical value proposition of deploying with Polygon CDK and orient this decision within the larger efforts of scaling Ethereum.

Initial Value Propositions Proposed by cLabs

  • Cross-community collaboration

From cLabs’s proposal: “While learning from proposals as they’ve been released, officially aligning to the Ethereum community offers the Celo ecosystem an opportunity to not only collaborate with the Ethereum community on existing EIPs, but also introduce our own, with near term examples being exploring standard for off-chain data availability and decentralized sequencers.”

Polygon Labs is eager to work alongside core Celo developers to standardize off-chain data availability and develop the best decentralized sequencing tech possible. As for off-chain data availability, Polygon CDK allows configurability for DA solutions, with the ability to integrate with third-party DA providers. As for sequencers, decentralized sequencing is on the roadmap for Polygon CDK, and Polygon Labs would love to work closely with cLabs developers to find solutions that meet the Celo community’s needs.

  • Enhanced compatibility with Ethereum

From the cLabs proposal: “Migrating Celo to utilize the OP stack eliminates the need to monitor compatibility, making it easy for Celo developers to utilize the full gambit of Ethereum tooling / libraries.”

cLabs understandably wants to maintain the network effects of Ethereum alignment, and the best way to maximize that is with EVM-equivalence. Polygon zkEVM is the only EVM-equivalent ZK rollup live on Ethereum Mainnet. EVM-equivalence means that an L2 built with Polygon CDK will be compatible with the EVM at the bytecode-level: any tooling and smart contract that works with the EVM, works with Polygon CDK. Relative to EVM-compatible implementations, EVM-equivalence provides greater security by reducing the surface area for potential bugs.

  • Increased security

cLabs has listed increased security as a value proposition for Celo becoming an L2 for Ethereum. Polygon CDK offers optimal security without sacrificing any other advantage of becoming an L2.

Any L2 chain deployed using Polygon CDK inherits the full security of Ethereum via ZK validity proofs. Every transaction is proven on Ethereum L1 with a mathematically verifiable proof, a process which happens quickly and doesn’t rely on economic incentives or any human action.

Polygon technology already has the fastest deterministic finality of any L2. And core Polygon developers are on track to build a type-1 zkEVM, which means the greatest possible compatibility–including in terms of security–with Ethereum while still retaining the scaling advantages of an L2.

Building on Polygon CDK would allow Celo to harness all of these benefits.

  • Maintain low fees

cLabs cites unchanged, low gas fees as essential in deploying an L2 validium. Not only does Polygon CDK offer an off-chain data availability solution for ZK-powered L2s running in validium mode, but by joining a shared ZK bridge, Celo developers and users would see fees that are magnitudes lower, as proof verification on L1 is amortized across all chains. Relative to the fees for transacting on Celo today, transaction fees as an Ethereum L2 built with Polygon CDK are anticipated to be on-par based on settlement time and transaction volume. The higher the transaction volume the cheaper the overall fees are due to amortization.

Value Propositions Specific to Polygon CDK

  • Unified Ecosystem of L2s

One of the biggest reasons to become an Ethereum L2 is to gain access to as much liquidity as possible. Polygon CDK gives a ZK-powered Celo L2 seamless access to deep liquidity.

Joining the Polygon ecosystem means access to over $23 billion in shared liquidity across all Polygon chains and Ethereum itself. Every chain in the Polygon ecosystem shares automatic access to the liquidity of the ecosystem–and of course, via ZK validity proofs, with Ethereum. As a result, users and developers can seamlessly transact or move across chains.

The deeper the pool of liquidity, the more opportunities for developers and users to harness and create value across all use cases.

  • Customizability and interoperability

In the above proposal to move to an L2, cLabs indicated a divergence from the OP stack in the following ways:

  • Decentralized sequencer
  • Off-chain data availability layer
  • A design that retains Celo’s 1-block finality

Point by point, Polygon CDK allows for out-of-the-box customization of each of the above parameters.

A ZK-powered L2 Celo chain could be deployed with a decentralized sequencer that uses existing Celo validators, widely configurable data availability solutions, and customization for block finality and the time it takes to post ZK proofs of the chain state to Ethereum. Such a ZK L2 would be natively interoperable with all other Polygon CDK chains. This means growing access to expanded Ethereum blockspace, all connected to a shared ZK bridge that allows near-instant cross-chain activity and access to shared liquidity.

  • Near-instant withdrawals and fast finality

ZK proofs offer the fastest deterministic finality of any rollup design, and, should Celo go forward with this proposal, its community would have access to the fastest finality among ZK rollups. Deterministic finality is how long it takes for a user to withdraw their funds on the L1. This is distinct from transaction finality, which is when a user sees that their balance has changed. With fraud proofs, reaching deterministic finality requires waiting the seven-day challenge window. Other ZK rollups live on Ethereum take at least 21 hours to reach deterministic finality. Currently with Polygon CDK, chains will experience 1 hour or less to reach deterministic finality. Current improvements are being made as we speak to improve finality times.

As it is today, Polygon CDK technology is orders of magnitudes faster; future optimizations to the prover are expected to reduce this even further.

The ability to reach deterministic finality this quickly is a reflection of the technical advantages of ZK proofs, generally, and the soundness of Polygon proving technology, specifically.

  • Support

In keeping with Celo’s stated desire to work more closely across the Ethereum community, Polygon Labs will provide a set of trusted implementation partners to assist Celo core devs to develop, test, and deploy a CDK Chain to best fit their required specifications. Similarly, Polygon Labs developers will aid in these conversations to ensure proper support is given. Polygon Labs developers have significant experience helping projects deploy L2s and would be eager to help.

Proposed Upgrade for Celo using Polygon CDK

The following diagram is one proposed design for Celo, in keeping with the stated requirements of:

  1. Decentralized sequencer
  2. Validium mode
  3. Third-party data availability solution

Additionally, future upgrades to the type of proving system (type-1, for instance) could be integrated into chain design.

Potential Effects on the Celo Community

In following with cLabs’ above proposal, Polygon Labs has identified the potential effects on various stakeholders across the Celo community.

  1. Validators

Change – The Polygon CDK offers L2s the ability to deploy in rollup or validium mode. Both configurations would allow for the participation of the current Celo validators, and would require validators to run or have access to a trusted full Ethereum node. In validium mode, validators would act as a data availability committee, responsible for maintenance and storage of transaction data. In rollup mode, Celo validators would act as sequencers. While decentralized sequencing is on the roadmap for Polygon CDK, current Celo validators would continue to play an important role in the health of the Celo ecosystem regardless of the implementation.

  1. Developers

Change – Developers will not experience any changes to what Celo currently offers, but rather see important improvements. Polygon CDK is EVM-equivalent, meaning any tooling (including auditing and debugging tools) that works with Ethereum, works with Polygon CDK. The upside for developers is at the heart of this proposal: an EVM-equivalent Ethereum L2 maximizes the network effects of Ethereum alignment. Meaningful steps have been taken to develop a forced-transaction mechanism from L2 to L1.

  1. Users

No change – Even in rollup mode, changes in transaction fees with Polygon CDK are expected to be relatively trivial. (As a point of reference, per-tx cost on Polygon zkEVM for an ETH transfer is ~$0.01.) By choosing to deploy in validium mode, a ZK-powered Celo L2 would see a significant reduction in fees. Furthermore, proof generation to Ethereum L1 is amortized across all ZK L2s deployed using Polygon CDK.

  1. Holders of CELO

No change – CELO will continue to be the token governing the Celo ecosystem. CELO holders will still control both the core smart contracts that control the protocols, according to Celo’s governance model, and the validator set. Additionally, CELO would continue to be the native gas token for transaction fees in the Celo ecosystem.


cLabs has highlighted a number of value propositions for upgrading to an Ethereum L2. Wth the expanded above proposal, Polygon Labs hopes that the Celo community will begin to discuss how deploying a ZK L2 using Polygon CDK can meet and surpass the original value propositions outlined by cLabs for a proposed Celo upgrade to an Ethereum L2.

We look forward to continuing this dialogue with the Celo community.


Thanks @SandeepN and welcome to the Celo Forum!

Looking forward to having us discuss this here in more detail with cLabs team and the broader community!


Hi @SandeepN, it is great to see you in the Celo forum! Thank you for taking the time to draft this thoughtful post.

Can you please help me understand how this works in practice?

Every chain in the Polygon ecosystem shares automatic access to the liquidity of the ecosystem–and of course, via ZK validity proofs, with Ethereum.

What does the interop roadmap look like?

What sort of compatibility, if any, does Polygon CDK have with the OP stack?

Is the CDK fully open source?

How do you anticipate this proposal would impact CELO owners, compared to how the OP proposal would impact CELO owners? That is, are there any considerations that might materially impact demand for CELO?

Thanks :pray:


Really interesting proposal, thanks for writing it up! I think I might be wondering something roughly similar to @Patrick insofar as how this relates to the existing cLabs L2 proposal: cLabs Proposal for Celo to transition to an Ethereum L2 - #60 by SandeepN

Would this represent a completely ground-up alternative to the previous plan? Or is there some common ground between them?

Can anyone comment on potential downsides associated with a change in direction here?


Hi Sandeep,

Welcome to the forum! It’s a real pleasure. This is a very long and detailed proposal. Thanks for spending the time on it and bringing it to the community.

I’m Tomer, a VC investing in the space and a Celo and Ethereum validator.

If I understand correctly, to meet one of the original goals of cLabs’ proposal, the network would have to operate in Validium mode under your proposal using Celo validators for data availability.

As an Ethereum validator, one of the things that excited me in the cLabs proposal was the use of EigenDA for data availability. Would this proposal be compatible with EigenDA?

Thank you!


Hi @SandeepN, thanks a lot for your forum post and all the work that went into it! I appreciate the well thought out overview of what Polygon has to offer with the Chain Development Kit, and how this can help to launch Ethereum L2s.

From the cLabs side, as one of the core technical contributors to the Celo blockchain, we’re open to consider different technical alternatives to reach our goals, namely to foster a financially inclusive community and bring real world use cases to Web3. We’re always happy to dive deeper into how to best reach those goals (technically).

We’re eager to learn more about what the broader Celo community thinks about this proposal, as these types of technical dives require time and effort and ultimately, it should be the community that decides how it wants Celo to evolve. Therefore, we invite everyone to share their thoughts, ask questions and express what dimensions they consider most important in choosing the tech stack with which to build Cel2.

Looking forward to an informative discussion here!


Welcome, @SandeepN; thanks for taking the initiative with this proposal and preparing such a focused write-up.

As a builder in the Celo space and part of Celo EU, I’ve been closely following the Celo L2 transition.

While others are better positioned to make the technical evaluations, I’m eager to explore co-building efforts with the ecosystem we would join.

I’m aware of The Value Prop, an excellent collaboration between Celo and Polygon, and eager to explore other ways you would imagine the ecosystems integrate and build together.

One of the elements that spoke to us a lot in the OP ecosystem was the Retrospective Public Goods Funding they run. It seems like a novel mechanism that rewards community value and innovation. As many Celo projects operate at the frontier of real-world development, I would imagine many of them to be eligible for RPGF in the case of Celo aligning with the OP stack.

Are there any active or future programs comparable with Retrospective Public Goods Funding or other avenues you would imagine Polygon and Celo supporting the development and growth of projects in the ecosystem?

I see a lot of value in joining forces on the non-technical side, from adoption and policy campaigns to events. Looking forward to more input and a potential collaboration.


Name is Vinit. I run strategic initiatives at Polygon. Sandeep is traveling back and forth so I am going to take a stab at some of these answers so we can keep the conversation going :slight_smile:

So first off, the Polygon protocol tech and architecture is moving to be completely ZK powered. Whether you’re on a rollup or validium mode using the CDK, the chain will have the ability to submit zk proofs to Ethereum via our architecture.

What does the interop roadmap look like?

The first step in the interop roadmap is attaching every CDK chain to one another via the LxLY bridge - the bridging solution for ZK chains in the Polygon ecosystem. Today the LxLy bridge only supports a 1:1 mapping (LxLy to CDK chain) so the work to enable 1:n is in progress and expected later this year. As of now, the LxLy bridge is up and operational. What this entails is that CDK chains are able to trustlessly get access to liquidity from Ethereum. They can also share liquidity effectively from one chain to another.

We’re also working on developing public posts on the next steps; for instance, on how each chain is going to reach our vision of Unified Liquidty. We’ll post those here once they are ready.

Compatibility with OP stack

Thank you for the thoughtful question. Can you please clarify what you mean by compatibility so I can better answer your question?

Is the CDK fully open source?

The code underlying CDK is open source similar to the approach taken with Polygon zkEVM. You can access the documentation to launch your own CDK: https://wiki.polygon.technology/docs/cdk/

Are there any considerations that might materially impact demand for CELO?

Let me review OP proposal once more and share thoughts on a future response. Thank you


Hey James great question. I referenced the initial proposal and wrote down some specifications the community wanted to see from their L2. I will list some of those requirement down low and provide answers to each one of them.

  • Decentralized sequencer powered by Celo’s existing validator set running Byzantine Fault Tolerance consensus.

With Polygon CDK, your validators will transition to act as a decentralized sequencer. Nothing has changed.

  • Off-chain data availability layer, powered by EigenLayer and EigenDA, operated by Ethereum node operators, and protected by restaked ETH. EigenDA brings the design of danksharding to Ethereum early, enabling Celo to maintain its low fees.

The CDK has two modes: Rollup and Validium mode. Validiums use off-chain Data Availability solutions such as a EigenDA, Avail, Celestia or another DAC. We can help you explore using da options that fit the specifications to keep fees low while having heightened security.

  • A design that retains Celo’s 1-block finality.

Transactions which are sequenced by the existing Celo validator set would retain the 1-block finality of transaction ordering. However, the finality of the bridge state is determined by the frequency of proofs being posted to L1.

  • Further Ethereum alignment and EVM compatibility, fostering a seamless developer experience

By connecting the Celo ecosystem to Ethereum using the Polygon CDK-driven proofs, Celo becomes strongly aligned with the Ethereum ecosystem.

  • Stronger security assurances than Celo provides individually

Polygon CDK chains will be able to prove their execution to Ethereum from day one, providing fast finality to L1. Contrary to optimistic rollups, this happens without relying on long and complex fraud proof mechanisms. The period of time where users must trust Celo validators for optimistic finality thus decreases dramatically.

  • A trustless bridge to Ethereum, simplifying liquidity sharing between Celo and Ethereum

This requirement is satisfied by the LxLy bridge system. Furthermore, because of the use of execution proofs the finality of bridge transactions can happen quickly.


Great question @Tomer-Ba, here some thoughts.

In addition to the Celo validators providing data availability as part of their role as sequencers, transaction data can also be posted to any other data-availability solution such as EigenDA, Celestia or Avail.


This is a great question. Polygon Labs recently announced the proposal to transition the Polygon ecosystem to Polygon 2.0, which supports a constellation of zk powered L2s that submits proofs to ETH. This proposal included 2% annual inflation for the POL token, of which 1% would be set aside for community/foundation level operations/incentivization. It will depend on the community, but we could see a % of these funds used to give public good funding to app developers to continue building innovative solutions.

Additionally, the Polygon community has pioneered finding sustainable methods to fund public goods that go beyond airdrops, for example the new L2 DAI bridge which leverages the DSR to give funds to Ethereum core development. The Celo community could use a similar solution

Regarding joining forces from an adoption and policy side, Polygon Labs is happy to explore.


Thanks for your reply @VinitParekh.

To clarify my question about the OP stack, could cLabs continue with their current roadmap to transition to the OP-Stack and later layer on Polygon CDK?

E.g. use the OP-Stack sequencer and batcher in combination with the Polygon CDK zk-proofs and bridge?

Also, did you find any considerations that might materially impact demand for CELO?


@SandeepN This is a compelling proposal.

Additionally, the work Polygon is doing in the global payments space integrating crypto and fiat with partnerships such as Stripe is commendable.

My inquiry revolves around public goods, and Polygon’s strategy for funding them in perpetuity. Could you share more with the Celo community around Polygon’s public goods funding strategy looking forward?


Thanks for this and all the other thoughtful answers - what’s the best way to follow or get involved in Polygon Community Governance from your POV? I would love to dip my toes into the latest and learn more about Polygon strategies, processes, and priorities ahead of a potential marriage :slight_smile:

1 Like

This sounds like a good proposition. I’m all for reducing duplicate efforts building the same solutions.
Can you elaborate on the maturity of CDK? any chains already using it?
Which features are still missing and what’s the timeline?

First, thanks @SandeepN for submitting Polygon’s proposal to help Celo’s transition to ETH L2. Second, I am thrilled with @matt & cLab/Celo’s mindset of opening to consider different technical alternatives.

It seems to me that Celo should maintain its competitive advantages - 1) decentralized sequencer 2) off-chain DA layer 3) 1-block finality. The component that is up to selection is the EVM execution. I believe that right open-minded approach to decide the best system for Celo L2 is to launch 3 testnets for OP EVM, Ploygon EVM, zkSync EVM. And request Celo projects to deploy and test each testnet end-to-end (e.g. not the just EVM execution but the whole L2 chain including decentralized sequencer, DA). Based on the results, select the best EVM execution and launch it as mainnet.

Evaluating Ploygon and zkSync proposals will unfortunately incur significant cost and additional work to cLabs and Celo projects. Plogygon and zkSync should provide financial and technical supports if they are serious about their proposals.

1 Like

Hey Celo team! Awesome to see that the community decided to make the transition to an L2 on Ethereum!

Given that this discussion resolves around Celo potentially using the Polygon CDK framework, I thought I quickly mention what we at Gelato have been working on as it would most likely enable the Celo team to make a decision about this topic more effectively.

We have recently launched our zkRollup as a Service platform which helps projects spin up & operate their own Polygon CDK L2s in no time and without having to deal with operating the underlying infrastructure (provers, rpcs, etc)! Check out this tweet for more info around our offering.

Astar Network which is Japanese leading blockchain originating from the Polkadot ecosystem, recently also made the decision to migrate to their own L2 based on the Polygon CDK framework and operated by Gelato (announcement).

We are more than happy to help the Celo team experiment with running a CDK testnet with minimal effort from their side to try out the technology. If the community decides that a Polygon CDK L2 is the way to go, we would of course also be more than happy to help set up & operate the production network with the desired configuration.

Feel free to reach out to us here or dm me on telegram (hilmarx) for more info! Happy to help if it’s useful!


Will Gelato be able to support Celo’s decentralized sequencer, off-chain DA and 1-block finality? If so, can you elaborate?

1 Like

Hi, yes Gelato’s Rollup platform is meant to support any configuration a project wishes to use, be it validiums with off-chain DA and potentially also decentralized sequencing once it becomes available. Happy to collaborate with the Celo team to determine what configuration would suit their needs best!


Have you worked with any off-chain DA and decentralized sequencer before? Or is it just in theory at the moment?