The blockchain industry is sixteen years old. In that time, roughly half a billion people have used one. Out of eight billion. That's a 6% penetration rate for a technology class that was supposed to restructure global finance. Email hit that number in four years. The web did it in seven.
The gap between what blockchains could be and what they actually are is not a mystery. It's a list: developers can't use their existing tools, users must buy a token before doing anything, institutions can't comply with securities law on a public ledger, and AI agents have no payment mechanism that doesn't involve a credit card form. These are engineering problems with engineering solutions.
Today, Casper Network publishes the Casper Manifest: a detailed technical roadmap covering specific protocol initiatives designed to close these gaps, targeting two converging markets that will define the next decade of blockchain adoption: regulated real-world asset tokenization and the emerging machine-to-machine economy. The Manifest comprises nine protocol initiatives — from EVM compatibility and networking hardening through gasless transactions, account abstraction, compliant security tokens, transaction privacy, micropayments, and quantum-safe cryptography — organized here in four thematic areas.
The last eighteen months were spent on the kind of work that never trends on crypto Twitter. In late 2024, the engineering organization was consolidated under one team with one roadmap. No more parallel workstreams shipping into the void. One prioritization stack. One release process. The results are measurable.
Casper 2.0, live since May 2025, was almost a full rebuild. It introduced the Zug consensus protocol with deterministic finality alongside a new multi-VM execution layer designed to support future virtual machines without architectural forks. The release also redesigned transaction payments and moved the protocol toward a unified model where user accounts and smart contracts are treated as the same kind of first-class object.
Casper 2.1 cut block time from 16 seconds to 8 and introduced protocol-level fee burning, permanently removing transaction fees from circulation. Validators still earn era rewards, but the supply now has deflationary pressure built directly into the protocol.
Casper 2.2 optimized tokenomics for long-term sustainability, and shipped in March 2026. A clean upgrade, delivered on schedule, passed through on-chain governance by the validator set after in-depth, ecosystem-wide debate.
Three protocol versions in under a year. The engineering machine is running. The question is what to build with it.
Before the roadmap itself, five design decisions already available in Casper’s protocol deserve attention. They were made during the lean years, when there was no roadmap to market them against. They are paying off now, because they make several of the Manifest initiatives structurally easier on Casper than they would be on any other chain.
Account and contract unification. While accounts and smart contracts look semantically different, they share much of the same underlying behavior. Both hold state and can have logic that changes that state. From the outside perspective, an account calling a contract is not fundamentally different from one contract calling another contract. Unifying accounts and contracts makes these interactions easier to compose and reuse, enabling richer behavior across the system. Most chains implementing account abstraction today are layering it onto systems where accounts and contracts remain fundamentally separate types. Casper instead moved toward unifying them at the architectural level — a decision that looks increasingly relevant as AI agents begin interacting with blockchains alongside humans, blurring the boundary between human users and machines. Smart accounts, session keys, and programmable spending limits are therefore extensions of the existing model rather than retrofits added later.
Deterministic finality. Zug consensus produces deterministic finality: once a block is confirmed, it cannot revert. Unlike probabilistic finality models, there is no expectation that additional confirmations are needed to achieve practical certainty. This matters for financial settlement, where transaction reversibility becomes a systemic risk. Ethereum still relies on extended confirmation windows for economic, probabilistic finality, while optimistic confirmation models such as Solana’s can revert under network stress. On Casper, all blocks are final.
Multi-VM support. Casper’s architecture allows different virtual machines to execute against the same global state. Today that means WASM. Tomorrow it can mean EVM alongside WASM. Adding support for another VM becomes an incremental extension, not a rewrite.
Fixed-cost operations. A native CSPR transfer costs exactly 0.1 CSPR. A delegation costs exactly 2.5 CSPR. Every operation is priced by a deterministic schedule enforced at the chainspec level rather than determined through dynamic fee auctions or fluctuating network demand. The same pricing model extends to new native operation types as they ship. When privacy operations land, their verification costs will remain deterministic as well. Casper will be the first general-purpose Layer 1 where a confidential transaction verification costs the same whether the network is at 10% capacity or 90%.
Cryptographic extensibility. Casper already supports multiple key and signature algorithms, including Ed25519 and Secp256k1. The architecture was designed so additional cryptographic schemes can be introduced without migrations, new account types, or breaking changes to the protocol. That matters for post-quantum cryptography, where many blockchains may eventually face disruptive transitions. On Casper, adding support for a new algorithm becomes an extension rather than a breaking change.
A Solidity developer has spent three years building on Ethereum. She knows the language inside out, has a library of audited contracts, and lives in Hardhat and Foundry. OpenZeppelin alone has over 27,000 GitHub stars. Casper has never been on her radar because it doesn't speak her language. MetaMask doesn't connect to it. Foundry doesn't deploy to it. The thousands of audited, production-tested contracts that protocols depend on don't run on it.
This is not a niche problem. The developer ecosystem runs on EVM tooling. Seventeen originally non-EVM Layer 1s added EVM compatibility in the last three years because ignoring this ecosystem means being invisible to it. Technical superiority has never been sufficient for adoption. VHS beat Betamax. x86 outlasted every RISC competitor for two decades. QWERTY is still here. Access matters more than elegance.
After the Manifest ships, that same developer opens MetaMask, connects to Casper, and deploys her contracts with the toolchain she already knows. Existing Solidity contracts work without modification. Her ERC-20 token operations and the existing WASM-side CEP-18 operations resolve to the same underlying protocol-level token state. No bridge. No fragmented liquidity. One set of tokens across two execution environments.
EVM Execution Engine. A full EVM using revm, the Rust implementation that powers Hardhat, Optimism, Coinbase and Foundry. This becomes the third VM in Casper's dispatch layer, executing Solidity bytecode against the same global state as WASM contracts. Same chain. Same finality. Same block. The dispatch architecture means this is an additive operation. EVM wallets, bridges, and indexers connect natively. Casper isn't becoming "an EVM chain." It's a polyglot chain that speaks EVM alongside everything else.
Networking Layer Hardening. Every other initiative in this document increases network load. EVM transactions are larger than WASM transactions. Post-quantum signatures are significantly larger than Ed25519. Batch transactions increase per-block payloads. Shipping those features requires a hardened and even more performant networking layer on a state-of-the-art stack.
Native Token Registry. User-defined tokens on Casper today (CEP-18, the ERC-20 equivalent) cost 4-8x more gas than native CSPR transfers. Every token is an isolated contract with no shared infrastructure. The Native Token Registry introduces protocol-level host functions that reduce costs by 3-5x immediately, then evolves into a full system contract alongside the existing Mint, bringing user tokens to CSPR-parity gas costs. In the long term, CSPR itself may become "currency zero" in the registry. One pricing model for all tokens. Fixed costs, chainspec-enforced. No other general-purpose Layer 1 offers deterministic-cost token operations at the protocol level. Additionally, the Native Token Registry will enable contracts built on any of the supported VMs (WASM, EVM) to interact with the same native tokens, and expose hooks for compliance and serialized functionality implementations to enable sophisticated regulated tokenization products.
A user in Jakarta opens a real estate investment app on her phone. She signs up with her Google account. She sees a property listing: fractional shares of a commercial building in Singapore, $50 minimum. She taps "Buy," confirms with Face ID, and owns tokenized equity in the fund. She didn't install a wallet. She didn't buy cryptocurrency for gas. She didn't sign anything she didn't understand. She doesn't know she's using a blockchain, and she shouldn't have to.
Under the surface, a stack of Casper Manifest initiatives made this possible. But the user doesn't care about the stack. She cares that it took one tap. That's the bar. And right now, no public blockchain meets it. Almost every transaction on every chain requires the user to already hold the native token. This is the equivalent of requiring someone to buy stock in Visa before swiping their credit card. Multi-step DeFi makes it worse: approve the token, wait, swap, wait, stake, wait. Three transactions, three signatures, three gas fees, three chances for the user to give up. Three chances for the business owner to lose a customer.
Gasless Transactions. Someone else pays your gas. Phase one uses off-chain EIP-2612/EIP-3009 permit signatures where a relayer submits on the user's behalf. This ships without protocol changes; the casper-eip-712 crate for typed data hashing already exists and works. Phase two brings fee delegation directly into the protocol through two mechanisms: a gas_payer field on transactions, and PricingMode::Prepaid, which is architecturally unique to Casper. Prepaid mode allows a sponsor to purchase execution receipts into escrow. The user references the receipt. No token balance required. Receipt-based gas prepayment at the protocol level is something Casper will be alone in offering among general-purpose Layer 1s.
Batch Transactions. Multiple operations in a single atomic transaction with configurable atomicity. Approve a token, swap it, and stake the result in one signature. If any step fails, everything reverts. No half-executed states. No three-transaction flows bleeding users at every step. DeFi starts to feel like a normal application. No need for developers to implement custom, external intent frameworks: this works natively on the Casper L1.
Smart Accounts. Because Casper 2.0 already moved toward a unified account and contract model at the data layer, “upgrading” a basic account into a smart account becomes an activation rather than a migration. Most chains implementing account abstraction today are adding it onto architectures where accounts and contracts remain fundamentally separate. Casper requires no ERC-4337 workarounds or proxy patterns. Instead, the underlying model supports programmable accounts directly at the protocol level. The result is protocol-native passkey authentication through WebAuthn, allowing users to sign blockchain transactions with familiar device biometrics instead of managing raw private keys, alongside session keys with time-bounded permissions, programmable spending limits, social recovery, and multi-factor authentication.
An asset manager wants to bring a $50 million commercial real estate fund on-chain. On most blockchains today, that looks like: deploy a token contract, build a separate compliance layer off-chain, and hope nothing falls through the cracks. The token itself is permissionless. Anyone can receive it. Freezing a bad actor requires a custom smart contract that may or may not work as advertised. Regulators look at this setup and walk away.
Simultaneously, a pension fund is evaluating blockchain platforms for a decade-long deployment. Their checklist eliminates most chains before the conversation starts: quantum resistance, on-chain compliance enforcement, transaction confidentiality, settlement finality, cost predictability. Most chains offer one or two. None currently offer all five.
The industry's answer so far has been to pick a side: compliance or privacy. Permissioned chains give you compliance and kill composability. Privacy chains give you confidentiality and make compliance impossible. No production chain currently offers both. This is the gap Casper fills.
Compliant Security Tokens. A full ERC-3643 equivalent for Casper: identity registries linked to on-chain claims, a pluggable compliance engine with modular rules (country restrictions, transfer limits, investor accreditation checks, supply caps), trusted issuer registries, and factory contracts for standardized deployment. ERC-3643 currently governs $28 billion in tokenized assets on Ethereum. Casper brings the same standard to a chain where the compliance engine evolves into protocol-level enforcement via Native Token Registry hooks, executing at native speed rather than smart contract gas costs. The fund's transfer restrictions follow the token, not the intermediary. Casper is taking a leading role in expanding ERC-3643 as the global standard for compliant tokenization, and adding its special sauce on top.
Beyond the Manifest's current scope, this infrastructure opens the path to a compliant on-chain orderbook: a central limit order book where every participant is identity-verified and every trade clears against the compliance engine. T+0 settlement with KYC'd counterparties — made possible by Zug's deterministic, single-block finality, which eliminates the confirmation windows that make real-time clearing impossible on probabilistic chains. Jurisdictional controls at the protocol level. No custodian. No T+2 delay. That's what institutional DeFi actually looks like when you build it on the right foundation.
Transaction Privacy. A multi-phase roadmap, ranging from stealth addresses for receiver privacy, all the way through private smart contract execution. The critical insight: by implementing privacy primitives as native host functions, Casper gives them fixed, chainspec-defined costs. No other general-purpose Layer 1 can offer deterministic-cost private transactions. And the system is designed to be compliant from the start. Viewing keys let auditors verify holdings without seeing transaction details. Proof of Innocence handles sanctions screening. Selective disclosure gives regulators what they need without exposing the fund's trading activity to competitors. Privacy and compliance as two sides of the same system, not opposites. This is the position no other chain occupies.
Quantum Safety. Casper’s cryptographic key system was designed for exactly this kind of extension: adding quantum-safe algorithms such as ML-DSA-44 or FIPS 2026 as new supported cryptographic schemes. Casper will ship hybrid accounts that carry both classical and post-quantum keys during a transition period, alongside migration tooling for existing accounts.
No major smart contract platform has shipped post-quantum transaction signing. The "harvest now, decrypt later" threat is real: an account you create today can have its signature broken by adversaries when quantum hardware matures. Enterprise clients evaluating blockchain platforms for ten-year deployments are starting to ask this question. Casper's answer will be production code, not a research paper.
Native Token Infrastructure. The Native Token Registry is the backbone that makes everything above practical at scale. It brings user-created tokens to CSPR-parity gas costs, provides the protocol-level hooks that compliance engines use for transfer restriction enforcement, and supplies the token state that privacy operations conceal amounts within. With it, every token operation has a fixed, predictable cost, and a common, extensible feature set. Eventually, CSPR itself becomes "currency zero": no structural distinction between the native token and any other token on the chain.
A logistics company runs a fleet optimization service. Third-party AI agents need to query its routing API thousands of times per day. Today, that means API keys, subscription tiers, invoicing, and a billing department that processes payments 30 days after the fact. An AI agent cannot sign a subscription agreement. It cannot submit a purchase order. It cannot call a SaaS API that requires a monthly billing plan negotiated by a salesperson over three emails and a Zoom call. What agents need is per-request payment: I call your API, you want $0.003, I sign a payment authorization, you verify it, you serve the response. Done.
HTTP status code 402 was reserved for exactly this in 1997. It took twenty-nine years for the payment infrastructure to catch up.
After the Casper Manifest starts shipping: an AI procurement agent at a retail company needs real-time container location data from that logistics API. The agent sends a GET request. The server responds with 402 and a price: 0.05 USDC per query. The agent checks its allowance, signs an EIP-712 payment authorization, resubmits. The server verifies, settles, returns the data. Elapsed time: 400 milliseconds. No API key registration. No contract negotiation. No billing department. The agent's owner adjusts the daily spending cap from her phone.
X402 Micropayments. The full HTTP 402 payment flow: status code, payment negotiation, stablecoin settlement. The proof of concept is already complete and demonstrated end-to-end. EIP-712 was ported to Casper earlier this year and is being used already to provide the cryptographic foundation. Production hardening and facilitator integration are weeks away from completion. Casper will be the first WASM-native Layer 1 with production X402 support.
Agent Infrastructure. AI agents don't need their own technology stack. They need the same infrastructure humans use, configured for autonomous operation. Smart accounts with scoped permissions: spend up to $100/day on API calls, only on whitelisted contract addresses, never touch the account's other assets. Gasless operation through the same PricingMode::Prepaid system that serves human users. Verifiable identity through the same compliance framework built for security tokens. Every piece of this lives somewhere else in the Casper Manifest. When you build infrastructure with autonomous actors in mind, agent support falls out naturally from the rest of the architecture.
The Casper Manifest's nine initiatives:
Shipping them simultaneously is not realistic, and shipping them in the wrong order wastes effort. The work is organized in three tiers based on dependency chains, competitive urgency, and execution readiness.
2026 H2. These form the foundation and address the most urgent competitive gaps. Networking hardening. EVM execution engine (protocol-level production target by end of 2026). X402 production deployment (PoC complete, production hardening underway). ERC-3643 Phase 1 (identity registry and compliance engine, buildable today with no protocol changes). Items with no protocol team dependencies run in parallel from day one.
2026 H2 through 2027 H1. These build on the Tier 1 foundation and unlock the user experience layer. Gasless transaction permits (the cryptographic tooling already exists). Batch transactions and account abstraction. Native Token Registry Phase 1.
2027. Strategically important but dependent on earlier tiers. Transaction privacy. Post-quantum signing. Continued maturation of the Native Token Registry, networking layer, and gasless infrastructure.
These initiatives are not isolated features. They share infrastructure and make each other stronger as they ship.
The Native Token Registry provides the protocol-level backbone for compliance hooks, privacy operations, and cross-VM token unification. Gasless transactions and account abstraction share the contract and account unification backbone, and the same design pipeline for fee delegation. X402 reuses the EIP-712 signing infrastructure built for gasless permits and benefits from smart accounts for agent-operated wallets. The networking upgrades absorb the throughput demands of everything else: EVM compatibility, post-quantum signatures, batch transactions, micropayments. Privacy combined with compliance creates a capability that exists nowhere else in the industry.
The engineering organization shipped three protocol versions in the last eleven months. The cadence is established. Formal protocol enhancement proposals (CEPs) for each initiative in the Casper Manifest will be published. Every architectural choice traces back to the six design decisions already live on mainnet.
The chains that close the gap between half a billion users and eight billion will be the ones that make blockchain invisible to regular people, trustworthy for institutional capital, and natively useful for machines. The Casper Manifest is the plan to be one of those chains. Not a theory. A construction plan, with the strong foundation already poured.
The past is there to learn from. The future is ours to build.
.png)
