Introduction & Thesis Overview

A Transition Redefining Industry Standards: Verifiable Compute

Scalability has long been the defining constraint of decentralized infrastructure.

At the blockchain level, networks like Ethereum guarantee integrity by design with every validator re-executing every transaction to reach consensus. This redundancy ensures trust but creates a throughput bottleneck: as networks scale, duplicated computation becomes exponentially expensive.

At the smart contract level, this limitation translates into restricted functionality. Because every line of logic must be re-executed and verified by all nodes, gas costs rise linearly with computational complexity. As a result, smart contracts cannot efficiently access historical data, process large datasets, or perform advanced off-chain logic such as AI inference. Essentially, the very mechanism that guarantees security also limits expressiveness.

Current and Future Projection of Web3 Development - Source: 0xCheeezzyyyy, MementoResearch

Verifiable compute resolves this architectural trade-off by separating execution from verification. Computations occur off-chain, where they are fast and inexpensive, while zero-knowledge (ZK) proofs attest to correctness and are verified on-chain. This allows smart contracts to trust external execution without re-running it by combining scalability with cryptographic assurance.

This principle extends beyond blockchains. In the broader digital economy, AI systems and data infrastructures now determine decisions across finance, healthcare, and logistics — yet operate as closed, opaque silos. Whether centralized or decentralized, the same issue persists: computation has outpaced our ability to verify it.

Verifiable compute restores that equilibrium. By proving correctness rather than assuming it, it transforms trust from a social construct into a mathematical one that enables scalable, data-rich, and intelligent systems built on proof instead of promise. As Ethereum and other L1/L2s adopt zkVM-based architectures, verifiable compute is set to become the foundation of scalability itself, where trust, performance, and cost-efficiency advance together.

The Brevis Thesis: The Infinite Compute Layer

Brevis is a verifiable compute infrastructure that allows any blockchain or application to prove complex off-chain computation with on-chain finality. It transforms blockchain networks into scalable verification layers that can support financial, analytical, and AI-driven logic alike.

For example, a DeFi protocol could use Brevis to prove an off-chain risk model that evaluates positions across multiple chains, then verify the result on Ethereum with a single proof. No re-execution, no data exposure, and no loss of trust.

Built on a modular zkVM architecture, Brevis merges cloud-grade performance with blockchain-grade security. Its composable design lets developers plug in custom coprocessors for specific workloads that extends verifiability into any domain where correctness, provenance, or compliance matters.

Aligned with Ethereum’s scaling roadmap and the broader shift toward proof-based computation, Brevis reframes scalability as an economic primitive rather than a design constraint. Positioned at the intersection of verifiable computing, DeFi, and AI, it serves as the foundational layer for a trustless, scalable, and provably correct digital economy.

Report Overview

As AI, DeFi, and data economies converge, Brevis emerges as the foundational infrastructure for scalable, trustless computation for workflows that are both verifiable and boundless in scale.

This report will cover:

• Brevis’ Infrastructure for Infinite, Programmable Compute: A concise look at Brevis’s modular architecture powering scalable, efficient, and secure computation across DeFi, AI, and enterprise systems.

• Strategic Positioning & Network Advantage: How Brevis’ traction, performance, and ecosystem depth establish it as the default layer of the verifiability economy.

Brevis: Infrastructure for Infinite, Programmable Compute

Core Architectural Underpinnings

Brevis redefines how computation is executed and verified across decentralized systems by establishing a unified framework where off-chain performance and on-chain trust operate as one continuous pipeline. By abstracting away complexity, Brevis turns ZK technology into a universal, production-ready developer primitive that is ****accessible, composable, and effortless to integrate.

Extending Ethereum’s cryptoeconomic security into a modular, multi-service compute layer, Brevis dissolves the long-standing trade-off between scalability and decentralisation, creating a new paradigm where performance and trust coexist in equilibrium.

Brevis’s Key Architectural Traits

Brevis’s architecture is guided by five foundational principles that collectively enable a system designed for extensibility, efficiency, and inclusivity across diverse computational workloads.

• Modularity: Built from independent, interchangeable components that can be reconfigured to match specific application requirements. Developers can compose proof pipelines for distinct workloads without reengineering core logic.

• Flexibility: Supports multiple proving backends and customisable pipelines, giving developers full control over proof generation parameters, verification depth, and hardware optimisation.

• Extensibility: Designed for seamless integration with specialised circuits and acceleration modules, enabling new capabilities to be added without disrupting the underlying architecture.

• Performance: Engineered for high throughput and low latency, delivering industry-leading proof generation speeds on commodity hardware through optimised circuit reuse and aggregation efficiency.

• Future-Readiness: Continuously adaptable to emerging cryptographic standards and hardware advancements, ensuring long-term relevance as ZK technology and GPU performance evolve.

Together, these principles ensure that Brevis remains not just a product of current ZK research, but an infrastructure standard adaptable to new computational paradigms, including AI inference verification and cross-domain data integrity.

The Core Engine: Pico zkVM

The foundation of Brevis’s compute layer is Pico: a high-performance zkVM that unites execution and verification to enable any computation to be provable, programmable, and economically scalable. It serves as the execution layer for verifiable logic (from DeFi algorithms to machine learning) designed to make complex off-chain workloads provable on-chain.

At its core, Pico features a hybrid architecture that fuses a general-purpose zkVM with application-specific coprocessors. This dual design enables Pico to efficiently execute broad computational workloads while accelerating specialized tasks such as cryptographic primitives, state transitions, or data proofs. Developers can seamlessly extend Pico’s capabilities by adding custom coprocessors without sacrificing performance, interoperability, or composability.

Pico’s architecture supports two primary modes of integration:

Function-Level Integration (Precompiles)

Pico incorporates dedicated circuits for core cryptographic operations such as elliptic curve arithmetic, hashing, and signature verification. These precompiles offload resource-intensive primitives into optimized modules, dramatically improving efficiency and lowering proof generation costs. This foundational layer provides a secure, production-tested base for high-frequency applications like rollups, privacy-preserving transfers, and Layer-2 scaling where proof generation speed and reliability are paramount.

Application-Level Integration (Coprocessors)

Above this, Brevis integrates domain-specific computation modules, extending Pico’s functionality beyond basic primitives to support complex, verifiable logic. These coprocessors act as specialized computational extensions, built to handle domain workloads such as:

• zk Data Coprocessor: Enabling cryptographically proven analysis of historical blockchain data, effectively giving smart contracts memory, context, and intelligence.

• zkML: Supporting verifiable machine learning workflows by proving model training and inference correctness without exposing proprietary data.

Together, these two integration layers deliver flexibility and scalability within a single verifiable compute framework as the core foundation of Brevis’s infrastructure.

Pico Prism: The Multi-GPU Proving Breakthrough

Building on this foundation, Brevis extends its performance frontier through Pico Prism: a multi-GPU clustering framework that redefines real-world ZK scalability. Engineered for massive parallelism and efficiency, Pico Prism transforms proof generation from a research bottleneck into production-grade infrastructure, capable of handling Ethereum-scale workloads with unprecedented speed and cost efficiency.

Brevis Pico Prism Performance & Comparison - Source: Brevis

Under production conditions, Pico Prism achieved:

• 99.6% real-time proving coverage under 12 seconds for current 45 M-gas Ethereum blocks (with 96.8% coverage under 10 seconds), using only 64 × RTX 5090 GPUs at a total hardware cost of $128K.

• On 36 M-gas blocks, it averages 6.04 seconds per proof, and 6.9 seconds on full 45 M-gas workloads.

Compared to the previous leading prover, SP1 Hypercube, Pico Prism demonstrates major leaps in performance and efficiency:

These results mark a pivotal milestone, where it successfully became the first zkVM to meet the Ethereum Foundation’s Real-Time Proving (RTP) benchmark: to achieve over 99% block coverage in under 10 seconds. By halving both latency and hardware cost, Pico Prism makes real-time verification not only possible but economically sustainable.

ZK Data Coprocessor: The Specialized Blockchain Intelligence Layer

At the frontier of the Brevis architecture lies the ZK Data Coprocessor: a specialised, app-level module within the modular zkVM stack that transforms how smart contracts interact with blockchain data. Purpose-built to tackle one of the most persistent limitations in decentralised systems, this gives smart contracts the ability to access, analyze, and verify historical on-chain data with cryptographic certainty.

The Problem: Smart Contracts Are Blind

Traditional smart contracts, while trustless, operate in isolation. This means they are unable to recall past transactions, analyse behavioral patterns, or process cross-chain information without relying on external intermediaries or incurring prohibitive costs.

For example, a DEX wishing to offer loyalty discounts based on historical trading volume, or a protocol aiming to reward users for long-term participation, would face gas costs in the tens of thousands of dollars for on-chain computation. This makes such applications impractical and economically unviable.

The Solution: Verifiable Off-Chain Intelligence

Brevis App Workflow Outline - Source:Brevis Documentation

The ZK Data Coprocessor solves this by retrieving and analysing any blockchain data off-chain, where computation is efficient, and then generating a ZK proof that attests to both:

1. The correctness of the computation.

2. The authenticity of the underlying historical data on-chain.

This proof is then fed back to the smart contract, which can verify the result trustlessly, enabling access to memory, context, and intelligence without compromising decentralisation. In essence, the ZK Data Coprocessor transforms contracts from static executors of logic into adaptive, data-aware agents capable of processing time-weighted, historical, and even cross-chain information securely.

The Impact: From Data Access to Programmable Intelligence

By bridging the gap between blockchain data and verifiable computation, the ZK Data Coprocessor unlocks an entirely new class of applications once deemed impossible on-chain:

• Loyalty & Incentive Programs: PancakeSwap now offers tiered trading discounts based on users’ verified historical volumes.

• Reward Distribution: Euler allocates lending rewards proportionally to time-weighted deposit positions, proven via ZK proofs.

• Cross-Protocol Rewards: Usual runs over $300M in annual rewards across eight protocols through trustless, ZK-verified computations.

• Network Incentives: Linea distributed $1B LINEA tokens to users who could cryptographically prove their past contributions and on-chain actions.

A New Paradigm for Smart Contract Design

The ZK Data Coprocessor effectively eliminates the cognitive and computational blindness of smart contracts, giving rise to applications that are not only economically viable but also cryptographically fair. From dynamic user scoring and personalised on-chain experiences to cross-chain governance and data-driven DeFi analytics, Brevis transforms the blockchain into an intelligent execution environment.

All of these components converge into a single, cohesive architecture that defines Brevis’s structural superiority across the ZK space. What Brevis enables is a vertically integrated, end-to-end verifiable compute stack that serves as a superior complete computational substrate across different domains that has been proven in production.

ZKVM and Prover (1x 4090) Landscape Overview - Source:ETHProofs

Its key achievements underscore this distinction:

• Pico Prism (multi-GPU Pico) achieved near real-time proving on Ethereum L1 (~96.8% of 45M-gas blocks proved in <10s, and ~99.6% in <12s), with an average proving time of ~6.9s, running on 64 RTX 5090s (consumer-grade). This redefined performance economics, halving both latency and hardware cost versus peers like SP1 Hypercube delivering industrial-grade scalability with 68% fewer GPUs.

• ZK Data Coprocessor extended the very definition of smart contracts, evolving them from isolated executors into data-aware, context-intelligent agents capable of verifiable reasoning across time and chains.

Together, these breakthroughs position Brevis at the forefront of the proving landscape as it demonstrates true real-time performance, modular verifiable intelligence, and economic sustainability within a single ecosystem.

The Architectural Differentiation: Developer Inclusivity

Brevis’s architectural philosophy centers on universality through accessibility. Its dual-layer design via merging a general-purpose zkVM (Pico) with a suite of domain-specific coprocessors strikes an equilibrium between broad computational flexibility and deep performance optimization. This structure ensures that Brevis can scale across heterogeneous workloads without fragmenting developer experience or compromising proof efficiency.

Brevis’s Developer Inclusivity Impact Scale Outline - 0xCheeezzyyyy, MementoResearch

Yet Brevis’s true differentiation lies in extending this inclusivity beyond architecture into developer enablement. By abstracting away cryptographic complexity and offering familiar toolchains, Brevis transforms verifiable computation from a specialised niche into a standard capability accessible to all builders.

• Familiar Tooling, Zero Cryptographic Overhead: Developers can build in Rust and interact with the Brevis stack using conventional frameworks. All low-level proof logic such as circuit construction, aggregation, verification, and proof recursion is fully automated. This removes the need for ZK expertise while maintaining production-grade reliability.

  
  

• Simplified Integration and Rapid Onboarding: The Brevis AVS DevKit and production SDK provide pre-built primitives for proof orchestration, attestation verification, and workload monitoring. These modular libraries allow developers to embed verifiability in existing pipelines without rearchitecting their codebase or learning new languages.

  
  

• Operational Visibility: Brevis introduces a native observability layer that tracks the full proof lifecycle (off-chain execution → on-chain verification) with performance, latency, and cost metrics presented through real-time dashboards. This transparency transforms cryptographic assurance into auditable operational confidence.

  
  

• Stack-Agnostic Modularity: Designed to operate across any application stack, Brevis supports both on-chain and off-chain integration. Whether validating AI inference outputs, verifying historical DeFi data, or enforcing compliance logic in enterprise workflows, its modular design adapts seamlessly embedding trust at the computation layer rather than the application edge.

Through this inclusive architecture, Brevis redefines verifiability as a universal development primitive: something that can be adopted incrementally, integrated effortlessly, and scaled transparently. This accessibility is central to mass adoption where it allows proofs to become not an exotic add-on, but a native layer of assurance woven into the logic of modern digital systems.

Recognising The Strategic Impact

Brevis’s modular architecture and performance breakthroughs collectively establish it as the core infrastructure for verifiable computation at scale. By integrating GPU-accelerated proving (Pico Prism), programmable coprocessors (zk Data coprocessor), and developer-first abstraction (Rust-based SDKs), Brevis closes the gap between the cryptographic rigour of blockchain and the flexibility of modern compute infrastructure. This synthesis yields three defining strategic outcomes:

Unification of Scalability, Security, and Usability

Brevis dissolves the traditional trade-offs that have constrained decentralised systems. Computations once deemed impractical on-chain (due to gas costs, latency, or data volume) are now executed off-chain at high speed and low cost, yet remain provably correct. In doing so, Brevis extends blockchain-grade integrity to workloads once confined to centralized trust models.

Institutionalisation of Verifiable Compute

By standardising proof generation and verification across diverse workloads, Brevis creates a reusable trust substrate for the digital economy. Its proof infrastructure can underpin everything from DeFi reward logic to AI compliance auditing, forming a shared computational backbone akin to cloud infrastructure but cryptographically verifiable.

Expansion of the Verifiable Economy

As Brevis’s ecosystem matures, proofs become a new form of digital asset that is transferable, auditable, and composable across applications. This enables secondary markets for verified data, inference attestations, and computation results, creating new economic primitives around computational trust.

Ultimately, Brevis transcends the definition of a ZK framework that enables a computational trust layer for the next generation of decentralised and data-driven systems. By rendering computation infinitely scalable, mathematically provable, and universally programmable, Brevis establishes the foundational infrastructure for a verifiable economy where every operation, across every domain, can be executed with absolute certainty.

Innovation Discovery: Differentiated Niche with Unparalleled Distribution

A Categorical Establishment: An Early-Mover Advantage

Brevis has emerged as a category-defining ZK infrastructure, uniting commercial traction, technical leadership, and ecosystem scale into a single defensible position. Entering the market before ‘verifiable computation’ was formally recognised as a category, Brevis achieved product–market fit ahead of the curve by transforming ZK computation from research concept into a production-proven developer primitive. This early establishment created a durable distribution moat, where technical performance and ecosystem adoption reinforce one another to drive sustained network dominance.

Unlike zkVM frameworks that remain largely experimental, Brevis has demonstrated real, repeatable throughput across live production systems.

A Strong Ecosystem Momentum

Notably, Brevis stands as the only zk-native infrastructure to achieve true production adoption running on live, economically active ecosystems. Its multi-dimensional traction underscores both technological depth and market relevance.

Brevis Ecosystem Overview & dApps Landscape - Source:Brevis

• Production-Grade Adoption: Live deployments across DeFi, data, and infrastructure protocols.

• Scale in Numbers: 124M+ proofs generated, 94K unique users, and 6 supported blockchains.

• Ecosystem Depth: 20+ integrations with leading protocols including PancakeSwap, Euler, Linea, MetaMask, and Bedrock.

• Economic Impact: $223M+ in verified rewards and over $2.8B in TVL growth enabled through Brevis-powered incentive systems.

Proof of adoption at economic scale is the ultimate validation of infrastructure maturity and a benchmark few ZK projects have achieved. Brevis distinguishes itself not through theoretical performance but through verifiable impact in production, where its compute layer powers real transaction flows and incentive mechanisms across live ecosystems.

Notably, Brevis’s verifiable compute layer has been embedded directly into Uniswap v4’s routing infrastructure, accelerating v4 Hook adoption, deepening liquidity, and improving trade execution efficiency. This integration is backed by a development grant from the Uniswap Foundation, enabling Brevis to build the Trustless Routing Rebate Program — a ZK-powered incentive framework allocating up to $9 million in gas rebates to routers integrating Uniswap v4 Hook pools.

Brevis’s Strategic Differentiation

Beyond its proven production utility, Brevis’s strategic edge is defined by three reinforcing pillars:

1. Performance Leadership: First to meet Ethereum’s RTP standards, proving production-grade scalability and setting a verifiable baseline for future zkVMs.

2. Inclusive Architecture: Its modular design abstracts cryptographic complexity, enabling any application to integrate verifiable computation without specialised expertise.

3. Validated Commercial Model: Live across multiple ecosystems, Brevis powers high-frequency reward systems, proof-based loyalty programs, and cross-chain attestations, demonstrating economic viability rather than theoretical potential.

This scale validates Brevis’s differentiated position as the only verifiable compute layer with measurable economic throughput, purposefully transforming ZK computation from concept to infrastructure standard. These attributes converge into a structural advantage: a system that delivers both developer inclusivity and institutional-grade assurance, scaling cryptographic trust into a generalised economic layer.

The Culmination: A Reinforcing Distribution Flywheel

All this adds up to a distribution flywheel where Brevis compounds growth across both technical and economic vectors. Each layer of adoption strengthens the next, creating a reinforcing cycle that expands Brevis’s reach while improving its cost structure and performance efficiency.

The Brevis Adoption Moat Flywheel - Source:0xCheeezzyyyy,MementoResearch

1. Superior Technical Offerings → Enhanced Strategic Distribution

Brevis’s best-in-class technical architecture establishes it as the default verifiable compute infrastructure across DeFi, AI, and enterprise domains. Its modular, composable design lowers integration barriers for developers, enabling seamless adoption across chains and verticals. This technical edge translates into organic go-to-market traction, where superior performance, developer accessibility, and production-grade reliability naturally drive distribution across diverse ecosystems.

2. Broader Distribution → Compounded Accreditations

Each new integration amplifies Brevis’ visibility, expanding ecosystem surface area and deepening network entrenchment. As usage scales, verifiable proof generation across multiple domains builds quantifiable trust metrics.

These milestones reinforce institutional confidence, formalising industry-wide recognition and credibility. In turn, ecosystem participation by leading protocols, developers, and enterprises accelerates Brevis’s transformation from a product to an industry standard for verifiable computation.

3. Increased Credibility → Reinforcement of Adoption Moat

Rising credibility compounds into a durable adoption moat. As partners, investors, and protocols align around Brevis’s ecosystem, the platform benefits from network-validated legitimacy, making it the go-to verifiable compute layer for mission-critical applications.

This accumulated trust not only strengthens developer preference and brand authority but also raises switching costs, as integrated workflows, proof dependencies, and ecosystem tooling become deeply intertwined with Brevis’ infrastructure.

4. Scaled Adoption → Continuous Technical Reinforcement

At scale, Brevis’s growing proof corpus and integration diversity feed directly back into technical refinement. Each new workload contributes telemetry for performance optimisation, circuit efficiency, and cost reduction, creating a continuous feedback loop that sharpens Brevis’s technological edge.

This self-reinforcing cycle (where adoption drives improvement and improvement fuels further adoption) ensures Brevis’s sustained leadership as the most performant, trusted, and economically scalable verifiable compute platform in the industry.

Team Expertise & Accreditations

The project’s foundation is equally strengthened by deep academic and institutional credibility. Founded by PhD cryptographers from UIUC, MIT, and UC Berkeley, Brevis’s team has contributed multiple peer-reviewed publications advancing modern ZK research and is formally recognised by the Ethereum Foundation as part of Ethereum’s scaling roadmap.

Strategic Investment & Confidence

Brevis’s trajectory is reinforced by deep institutional conviction and cross-ecosystem alignment.

Brevis Seed Round Participation Highlights - Source: Brevis

It is backed by leading global investors such as Polychain Capital, Binance Labs, IOSG, Nomad Capital, HashKey, and Bankless Ventures. This accredited participation underscores confidence in Brevis’s long-term commercialization and category-defining potential.

Beyond capital, Brevis’s network of strategic partners includes key stakeholders from Kyber, Babylon, Uniswap, Arbitrum, and Altlayer. This also reflects a shared commitment to advancing verifiable computation as a foundational Web3 primitive. Together, this coalition forms the backbone of Brevis’s distribution moat, uniting institutional trust, ecosystem reach, and technical credibility.

Through this alignment of capital, partnerships, and purpose, Brevis cements its position as the definitive trust layer for verifiable computation: the infrastructure powering the next era of scalable, transparent, and provable digital systems.

Closing Thoughts

Brevis’s technology represents a fundamental shift in how computation, trust, and value coexist in the digital age. In an increasingly complex and opaque world, Brevis reintroduces a simple but profound principle: trust should be based on proof.

Since its inception, Brevis has achieved what no other ZK project has through delivering production-scale verifiable compute across live ecosystems with measurable economic outcomes. By translating its technical lead into a compounding distribution advantage, Brevis has established a self-sustaining network effect that scales organically within a rapidly emerging category.

As verifiability becomes a core requirement for digital infrastructure, Brevis sits at the center of this shift. Its technology combines performance and transparency, allowing applications to prove their correctness and reliability by design.

Through this structural dynamic and its singular position within the ZK landscape, Brevis emerges as a defining infrastructure built through distribution and proof: The infinite compute layer powering the verifiable economy.

Authors: @0xCheeezzyyyy, Memento Research

This report was written in partnership with Brevis. This report has been prepared for informational purposes only. It does not constitute investment advice, financial advice, trading advice, or any other sort of advice, and you should not treat any of the report’s content as such.