Biconomy and the Ethereum Foundation published ERC-8211 on April 6, 2026, a contract-layer execution standard that allows multi-step DeFi transactions to resolve parameters at runtime rather than at signing. The proposal, co-developed under the Foundation's Improve UX track, is aimed squarely at ...
"When you have an output from something like a swap, you don't know how much that will be. Developers have to either hard code that or find another way for that output to be used as an input. What we've built lets developers just say: whatever the balance is of the user, just compose that with the next action." — Ahmed Al-Balaghi, Co-founder, Biconomy
Biconomy and the Ethereum Foundation published ERC-8211 on April 6, 2026, a contract-layer execution standard that allows multi-step DeFi transactions to resolve parameters at runtime rather than at signing. The proposal, co-developed under the Foundation's Improve UX track, is aimed squarely at autonomous agents that need to chain on-chain actions without pre-computing intermediate values. It lands at a moment when agent-driven transaction counts are rising sharply but dollar volumes remain thin, and when the economic question — who pays the gas, who captures the fees — is still unresolved.
ERC-8211 does not require a protocol fork. It is a contract encoding compatible with ERC-4337 account abstraction and ERC-7683 cross-chain intents. The standard introduces three primitives: fetchers that read live state, constraints that validate resolved values, and predicates that gate execution between calls. In practice, an agent can express "swap token A on Uniswap, then deposit whatever arrives into Aave" as a single atomic batch, rather than as two signed transactions or a bespoke smart contract.
The announcement arrives alongside data showing a gap between agent adoption and agent revenue. Daily active on-chain AI agents crossed 250,000 in early 2026, a roughly 400% increase year-on-year, according to industry trackers. Yet the most cited agent-payments protocol, Coinbase's x402, processes approximately $28,000 in daily on-chain volume against a claimed $600 million annualized figure that includes testing flows. The plumbing is being laid faster than the paying customers are arriving.
ERC-8211 was published on April 6, 2026, with an open-source reference implementation and a live demo. The authors listed on the specification are Mislav Javor, Filip Dujmušić, Filipp Makarov, and Venkatesh Rajendran, all current or former Biconomy engineers. The work originated from a 2025 workshop convened by the Ethereum Foundation's Improve UX initiative.
The three primitives are narrow by design. Fetchers declare how each parameter in a batch is obtained — as a literal, via a static call, or from an on-chain balance. Constraints specify conditions values must satisfy before execution proceeds. Predicates act as inline gates between steps. Barnabé Monnot, a research scientist at the Ethereum Foundation, framed the timing: "The agentic execution angle is new, but has imposed itself given the rapid developments of agents over the last three months."
The standard is account-agnostic. It does not require users or agents to migrate wallet stacks. It layers on top of existing account-abstraction infrastructure, which means it inherits the cost structure, bundler dependencies, and paymaster economics of ERC-4337 rather than displacing them.
The backdrop is familiar: stablecoin supply on Ethereum reached an all-time high of $180 billion in early April 2026, with USDT at $80.7 billion and USDC at $51.8 billion on the network. Ethereum retains roughly 60% of the total stablecoin market. Settlement dollars are on-chain in volume. What is less clear is whether agents have meaningful settlement demand to route through them.
Solana Foundation executives have publicly forecast that agents, bots and LLM-based wallets will drive "99.99% of all onchain transactions" within two years. That is a forecast, not a current reading. The current reading is that agent transactions dominate by count in certain categories — arbitrage, liquidity rebalancing, MEV — but remain a small share of identifiable end-user economic flows.
Webthreepedia's October 2025 analysis of blockchain economic value distribution estimated $13–14 billion in annual transparent on-chain revenue across the entire sector, against $86–113 billion in total funding when subsidies, token unlocks and venture capital are included. Any new execution layer needs to be evaluated against that denominator. ERC-8211 does not create revenue. It makes existing execution cheaper and more composable. Whether that translates into captured value depends on who operates the bundlers, who writes the agents, and who pays for the infrastructure.
The problem ERC-8211 targets is concrete. Current batch execution on Ethereum locks calldata at signing time. An agent that wants to execute "swap 1,000 USDC for ETH on Uniswap, then deposit 100% of received ETH into Aave" cannot sign the second transaction until the first has settled, because the received amount is unknown at signing. Workarounds include hard-coded estimates, slippage buffers that leave dust, bespoke router contracts, or two separate user operations.
Each workaround imposes a cost. Hard-coded amounts lose value to slippage. Router contracts require audits and deployment gas. Two user operations double the bundler overhead and break atomicity — if the second transaction fails, the first has already executed.
Al-Balaghi described the alternative: "That means you can create these really powerful flows without writing new smart contracts. You can just do it in TypeScript." The claim is that a strategy previously requiring a purpose-built Solidity contract can now be expressed as a configuration object. If the claim holds at scale, it reduces the fixed cost of launching new agent strategies and shifts more logic into off-chain tooling.
The technical bet is that runtime resolution, combined with inline predicates, turns a batch "into a program with embedded safety checks, not a hopeful script," in the language of the specification. This is a non-trivial distinction. A hopeful script fails silently or catastrophically when a dependency is wrong. A program with embedded checks reverts cleanly and predictably.
For agents running strategies across Uniswap, Aave, Curve, Compound, Balancer and similar venues, predictable revert semantics matter more than raw throughput. Most agent losses in 2024 and 2025 were not from slow execution but from edge cases where intermediate state differed from the values encoded at signing.
Whether ERC-8211 becomes the dominant format depends on adoption by three groups: wallet vendors who must support the new calldata format, bundlers who must simulate the fetchers and constraints correctly, and protocol integrators who must expose state in ways the fetchers can read cheaply. None of these are technically hard. All three are coordination problems, and coordination problems take time.
The honest reading of agent-payments demand in April 2026 is that the on-chain numbers do not yet justify the infrastructure build. Coinbase's x402 protocol, launched with Cloudflare and Stripe and positioned as the HTTP-native standard for agent payments, has processed over 119 million transactions on Base and 35 million on Solana as of March 2026. The daily on-chain volume, however, is approximately $28,000, according to CoinDesk analysis. The $600 million figure cited in marketing materials is annualized and includes testing and "gamed" transactions rather than third-party commerce.
On April 2, 2026, Coinbase, Cloudflare and Stripe announced a new foundation to advance agent payments. The x402 ecosystem is valued at approximately $7 billion across backed projects. The valuation-to-revenue ratio is extreme even by 2026 crypto standards.
ERC-8211 is not a payments protocol. It is an execution standard. But it shares the same customer set. If agent payments remain a forecast rather than a business, the execution standard will be used primarily by the same MEV and rebalancing bots that already drive most programmatic on-chain activity, not by a new category of consumer-facing agents. Industry estimates put AI-driven programmatic execution at roughly $2.5 trillion in TVL touched annually through liquidity pools and arbitrage, but that figure captures activity by existing non-agentic bots alongside newer LLM-driven agents and does not cleanly separate the two.
ERC-8211 is a clean piece of plumbing. It solves a real problem — the static calldata constraint — that has forced agent developers into awkward workarounds for two years. The technical design is narrow, composable with existing account-abstraction infrastructure, and does not require users to change anything about their wallets or signing flows. Backed by the Ethereum Foundation's Improve UX track, it has institutional support and will likely ship into production bundlers within the current release cycle.
The question is not whether it works. It is whether the customers it is built for exist in numbers large enough to matter. Current on-chain data shows agent activity dominated by existing MEV and arbitrage bots rather than by a new class of autonomous consumer agents. The agent-payments ecosystem has raised capital and shipped protocols faster than it has generated third-party revenue. ERC-8211 lowers the fixed cost of building agent strategies, which is useful but does not create demand for those strategies.
For now, the standard's most likely near-term users are the same sophisticated actors — market makers, MEV searchers, rebalancing bots — who already operate on-chain at scale. The retail-facing agent economy that the specification is rhetorically aimed at remains, in April 2026, more forecast than fact. The infrastructure is real. The customers are provisional.