While we were still debating who has stronger zkVM performance,
@SuccinctLabs and @boundless_xyz have already turned their attention to another more practical issue:
zk proof, how to run it? Who runs it? Where does it run? How to monetize it?
The real competition is not in 'how to generate proofs,' but in 'who generates the proofs.'
Succinct and Boundless have both submitted two reports:
A unified platform that schedules proof tasks.
A locally deployed service, embedded into the protocol ecosystem.
This is not just two solutions, but two paradigms that zk Infra is currently differentiating.
1. Boundless: A set of locally deployed zk service engines.
Boundless is not a chain, nor is it L2, but a zk market system that deploys local contracts on each main chain or L2.
Its basic philosophy is: wherever the application is, deploy the proof market on that chain.
This means it does not create a unified central market scheduling, but is more like 'building local verification modules for each protocol,' with a user experience close to #Uniswap or #Aave's contract calls.
Its core advantage lies in:
Complete native chain security (all verification/settlement occurs on the native chain).
Eliminates cross-chain funding, no need for additional bridging or gas tokens.
Supports local sequencing, fair bidding, anti-censorship properties.
zkVM + off-chain logic minimizes on-chain footprint (a proof request as low as 35k gas).
The applicable objects are very clear: #rollup, #restaking, governance, real-time proof demand systems, etc.
This is actually a zk plugin developed for chain-level protocols.
2. Succinct: Build a unified proof market across chains
Succinct's approach is completely different; its goal is not to 'serve a certain protocol,'
Instead, it aims to create a zk proof service platform that can connect everything, route everything, and schedule everything.
This system consists of three parts:
SP1 zkVM: RISC-V instruction set, supports Rust programming, no need to manually write circuits.
Proof Market: tasks are submitted by demanders, Prover nodes bid through collateral & auction.
Multi-chain support: task sources support Cosmos IBC, Polygon CDK, Celestia DA, etc.
For example, if I send a request that needs verification, Succinct will help you complete: bidding, assigning a prover, generating proof, on-chain verification to automatic settlement.
This system does not run on every chain but serves as a general service embedded within the multi-chain ecosystem.
It is like Stripe + Render in the zk field, providing integrated services of developer interfaces + computing power scheduling + payment systems.
3. What is the difference between Succinct and Boundless?
Many people think that Succinct and Boundless are just two different implementations of products:
One builds a platform, the other builds components; one is centralized, the other is decentralized.
But actually, their differences lie not only in how they do it but in who they want to become.
Succinct aims to build a unified zk service platform, like Stripe or AWS, becoming the 'proof hub' between chains.
Developers only need to access the API once to cross-chain schedule the prover network, automatically generate, verify, and settle zk proofs.
Boundless took a completely opposite path:
It chooses to deeply embed itself into each chain, turning zk services into built-in capabilities of the protocol. The proof market does not exist independently but is deployed as local modules on each chain within native contracts.
This design deeply binds it with the main chain's security, ordering, and fairness, making it more like a 'ZK plugin' rather than a 'ZK platform.'
The two also form a very distinct contrast in architecture:
Succinct tends to unify the management of scheduling, verification, and payments, adapting to various ecosystems;
Boundless believes the most reliable way is to return all verification work back to the application chain to handle, ensuring safety and censorship resistance from the source.
This difference also extends to the scalability path:
Succinct's design facilitates standardization, integration, and ecological network effects,
It hopes that 'the more people connect, the lower the cost and the higher the efficiency';
Boundless emphasizes local autonomy more, with service depth greater than service range,
Every deployment feels like it is tailor-made for a certain chain.
You can see them as two typical roles in zk Infra:
One attempts to connect all chains, creating a global zk market infrastructure;
One chooses to serve each chain, achieving excellence, closeness, and stability in the local context.
They indeed represent two completely different development paradigms of zk Infra—
One turns zk into a network service, while the other integrates zk into protocol structure.
4. Why do I have more faith in Succinct?
I'm not saying Boundless is bad. On the contrary, its architecture is clear and self-consistent, and it can even be the optimal solution when serving certain specific types of protocols.
But from the perspectives of platform potential, network effects, and ecological expansion ability, Succinct clearly has more groundbreaking imagination.
It is more like the Stripe of the zk field.
What Succinct does is turn zk proofs into standard services; you don't need to understand circuits or change protocol logic.
As long as a request is made, it can schedule the most suitable Prover to complete the task.
Like Stripe, complex infrastructure is hidden behind the API,
Truly lowering the barrier to using zk.
It has already started running.
Succinct is not just a PPT or a story.
Its SP1 zkVM has been open-sourced, and the Proof Market has been launched.
It has even provided services for cross-chain verification of Cosmos IBC's Eureka and Polygon CDK's Validium system. It is 'already achieved.'
It has a platform-level flywheel effect
The more users there are, the more Provers there are, the lower the price;
The more protocols are connected, the more stable the standards become, and the more mature the interfaces are.
This is a network effect that only a platform can have,
Rather than a local advantage that a single protocol component can form.
It is suitable for a modular future.
With the evolution of Rollup-as-a-Service, DA layers, and chain abstraction,
The future on-chain world will increasingly rely on unified, general, cross-chain service components.
The combination of Proof Market + zkVM built by Succinct,
This is precisely designed for such 'modular collaboration' scenarios.
Of course, Boundless also has its own moat - local deployment, extreme compression, and clear security.
But in a Web3 world that emphasizes standardization and composability, I prefer Succinct as the role of 'on-chain verification infrastructure.'
