Ethereum in 2026 is a different network than the one that launched in 2015 — and a different network than the one that completed the Merge in 2022. The rollup ecosystem now processes the bulk of user transactions. The base layer is settling tens of billions of dollars of tokenised real-world assets. ETH itself has evolved from gas token to productive capital. This is a complete look at where Ethereum stands today.

Quick read.

  • ETH trades in the $3,800-4,200 range with a market cap near $500B, second only to Bitcoin among crypto assets.
  • The rollup-centric roadmap has matured — L2 throughput exceeds mainnet by orders of magnitude on most measures.
  • Real-world asset tokenisation (BlackRock BUIDL, Franklin Templeton BENJI) has scaled to billions of dollars on Ethereum and its L2s.
  • ETH is increasingly held as productive capital — staked, restaked, used as collateral — rather than purely as gas.

1. Ethereum’s role in the on-chain economy

Ethereum is the settlement layer for everything on-chain that requires a credibly neutral, broadly accessible, robustly decentralised base. The major stablecoins issue on Ethereum. The largest tokenised treasury funds settle on Ethereum. The DeFi protocols that hold the deepest liquidity exist on Ethereum or its L2s. When someone bridges from a non-Ethereum chain to Ethereum, they are choosing the most institutionally-trusted base layer available in the cryptocurrency ecosystem.

This is the bet that Vitalik Buterin and the Ethereum Foundation made years ago: that decentralisation has a price worth paying, and that the broader cryptocurrency ecosystem would converge on a single base layer that is credibly neutral above all else. The bet has paid off in the institutional adoption that has materialised — BlackRock’s BUIDL fund, Franklin Templeton’s BENJI, JP Morgan’s Onyx tokenised money market funds, and the major dollar stablecoins all chose Ethereum specifically because it cleared their threat-model bar.

The trade-off is that the base layer is intentionally slow and expensive when measured against monolithic alternatives. A simple transfer on Ethereum mainnet costs cents to a few dollars depending on demand, with confirmation times measured in tens of seconds. This is acceptable for institutional settlement; it is unacceptable for consumer payments. The answer has been the rollup ecosystem.

Layer 2 networks — Arbitrum, Base, Optimism, zkSync, Linea, Scroll, Starknet, Polygon zkEVM, and dozens of others — process the bulk of user transactions while inheriting Ethereum’s security through cryptographic proofs. A user doing high-frequency trading or sending small payments operates on an L2. Settlement back to mainnet happens periodically, automatically, and at much lower cost than if every individual transaction were executed on mainnet.

2. Technology and architecture

Ethereum’s core protocol has evolved substantially since the Merge in September 2022. The transition from Proof-of-Work to Proof-of-Stake cut energy consumption by approximately 99.9% and introduced native staking yields. The subsequent Shanghai upgrade (April 2023) enabled validator withdrawals, completing the transition. The Cancun-Deneb (Dencun) upgrade in March 2024 introduced EIP-4844 (proto-danksharding), which dramatically reduced data costs for rollups.

The execution layer continues to be the Ethereum Virtual Machine (EVM). The EVM is the most widely-deployed smart-contract execution environment, supported by essentially every Ethereum-compatible chain and L2. While alternative VMs (the Solana SVM, the Move VM used by Aptos and Sui, various WASM-based environments) have specific advantages, the EVM’s network effects in tooling, developer mind-share, and security expertise remain unmatched.

Consensus is achieved through a hybrid of LMD-GHOST (block production) and Casper FFG (finality). Validators are required to lock 32 ETH and run consensus client software. Approximately one million validator entities operate the network as of 2026, securing a staked supply of roughly 34 million ETH (28% of total supply). The diversity of consensus clients (Prysm, Lighthouse, Teku, Nimbus, Lodestar) is a notable improvement from earlier years where one client dominated. Execution client diversity (Geth, Nethermind, Besu, Erigon, Reth) has also improved.

The roadmap has several key components. The Verge introduces Verkle trees for stateless clients, dramatically reducing the storage burden on validators. The Purge removes historical state from full nodes, simplifying long-term operation. The Splurge encompasses a series of smaller improvements (account abstraction, EOA upgrades, RANDAO improvements). Full danksharding — the longer-term scaling target that would allow Ethereum to provide vastly more data availability for rollups — is still in development but progressing.

EIP-7702 (smart accounts) shipped in 2024 and has been transformative for wallet UX. Externally-owned accounts can now execute code temporarily, enabling features like batch transactions, gas sponsorship, and improved signature schemes without users needing to migrate to entirely new account types. This has been a major UX win that has propagated through wallet software.

3. Tokenomics and monetary structure

Ethereum’s monetary policy is fundamentally different from Bitcoin’s. There is no hard cap on supply. New ETH is issued to validators as a staking reward — currently roughly 600,000 ETH per year. A portion of every transaction fee is burned via EIP-1559 — historically averaging similar amounts to issuance, resulting in net-neutral or net-deflationary supply across cycles.

The current ETH supply is approximately 120.5 million tokens. Cumulative burns since EIP-1559’s introduction (August 2021) exceed 4.5 million ETH. During periods of high network activity, the burn rate exceeds the issuance rate and ETH becomes net-deflationary. During periods of low activity (as in mid-2024 and parts of 2025), issuance modestly exceeds burns and ETH is slightly net-inflationary.

Staking participation has continued to grow. Approximately 28% of total ETH supply is staked, generating yield for validators of roughly 3-4% annually (declining slightly as more ETH gets staked). The staked ETH is held in a variety of arrangements: solo validators (32 ETH each), staking pools (Lido, Rocket Pool, Frax, others), and increasingly through restaking via EigenLayer.

EigenLayer’s restaking model has introduced a new dimension to ETH yield. Staked ETH (or LSTs like stETH) can be restaked to provide cryptoeconomic security for additional services (AVSs — Actively Validated Services). These services pay rewards on top of base staking yield. Total restaked ETH exceeds 5 million as of 2026, generating combined yields that can reach 5-7% for active restakers. The risk is meaningful — additional slashing conditions apply — but for many ETH holders the additional yield is attractive enough to participate.

Beyond staking and restaking, ETH is widely used as DeFi collateral. The aggregate ETH locked in DeFi (lending protocols, liquid staking, restaking, structured products) easily exceeds 30 million ETH. This represents a meaningful share of total supply that is structurally not available for trading liquidity — productive capital working across multiple protocols simultaneously.

4. Adoption and key metrics

Ethereum mainnet active addresses, transaction count, and fee revenue have all stabilised at levels meaningfully higher than pre-2022 but below the absolute peaks of late-2021. This reflects a structural change: most user activity has migrated to L2s. The mainnet activity that remains is concentrated in higher-value transactions — large stablecoin transfers, DeFi interactions involving meaningful capital, RWA tokenisation flows, and bridge transactions to and from L2s.

The L2 ecosystem in aggregate processes orders of magnitude more transactions than mainnet. Arbitrum, Base, Optimism, and the various ZK rollups (zkSync, Linea, Scroll, Starknet) collectively handle 10-20 million transactions per day, depending on activity. Total L2 TVL exceeds $40B. The user experience on most L2s is now competitive with monolithic L1 alternatives in terms of cost and speed.

Real-world asset tokenisation has become the institutional story for Ethereum. BlackRock’s BUIDL fund — a tokenised money-market fund issued on Ethereum — has scaled to billions of dollars. Franklin Templeton’s BENJI fund operates similarly. JP Morgan’s Onyx platform tokenises money-market funds on permissioned Ethereum-compatible chains. Major banks have begun issuing tokenised treasury products on Ethereum mainnet specifically because of regulatory comfort with the platform’s institutional adoption pattern.

Stablecoin issuance on Ethereum remains the dominant share of total stablecoin supply. USDC, USDT, DAI, FRAX, GHO, PYUSD, and others issue significant volumes on Ethereum mainnet. The aggregate stablecoin supply on Ethereum exceeds $130B as of 2026, of which a growing portion has migrated to L2s for active transactional use while remaining bridged to mainnet for settlement.

For current network metrics and ETH market data, see our Ethereum profile.

5. Risks

Staking centralisation. Lido alone controls approximately 28% of staked ETH. Coinbase, Binance, and other large staking pools collectively control much of the remainder. The 33% threshold for “concentration concern” remains a live issue. Ongoing work on distributed validator technology (DVT) and module diversification aims to mitigate this, but the structural pressure toward concentration remains.

L2 fragmentation. The proliferation of L2s has created a fragmented user experience. Users must hold ETH on multiple L2s, manage bridging between them, and understand which protocols are deployed where. Cross-rollup interoperability is a major roadmap item but has not yet shipped at a level that fully solves the UX problem. The risk is that users find alternative monolithic L1s (Solana, particularly) easier and migrate their activity.

Rollup centralisation. Most L2s today have centralised sequencers operated by their respective teams. Stage 1 decentralisation has been achieved on several L2s (Arbitrum, Optimism, Base via OP Stack), but full sequencer decentralisation remains a work in progress. Sequencer-level censorship or downtime affects every user on a given L2.

EigenLayer concentration. Restaking introduces a new attack surface — slashing conditions across multiple AVSs that compound risks for individual validators. The systemic risk of widespread slashing events affecting a large share of staked ETH is theoretical but real. Risk management practices in the restaking ecosystem are still being developed.

Execution complexity. The Ethereum codebase has grown substantially over the years. The Cancun-Deneb upgrade alone added thousands of lines of consensus and execution code. The risk of subtle bugs in client implementations, particularly during major upgrades, is non-trivial. The client diversity story partially mitigates this but does not eliminate it.

Regulatory shifts. ETH staking specifically has been the subject of US regulatory scrutiny. The eventual treatment of staking yields, restaking yields, and LSTs under federal securities law remains incompletely resolved. Adverse rulings could materially affect demand from regulated institutional holders.

Competitive pressure. Solana has demonstrated that monolithic chains can serve high-throughput consumer applications competitively. While Ethereum’s institutional positioning is strong, the application-layer competition for retail users, payments, and consumer-scale apps is ongoing and not guaranteed to favour the rollup-based approach indefinitely.

6. The model’s take

Our quantitative model produces a wide distribution of outcomes for ETH, similar in shape to Bitcoin but with somewhat higher implied volatility. The base case has ETH appreciating against current levels over multi-year horizons driven by continued L2 adoption, RWA tokenisation growth, and the deflationary mechanics of EIP-1559 during high-activity periods. Upside scenarios involve continued institutional adoption of Ethereum-based tokenised assets and continued growth in restaking yield capture. Downside scenarios involve material competitive loss to alternative L1s or adverse regulatory developments specific to staking.

The model cannot incorporate certain discrete events: a major sovereign or central bank issuing tokenised securities on Ethereum, a regulatory framework that explicitly favours or disfavours staking-based assets, or a technical breakthrough in scaling that changes the L2 versus monolithic-L1 calculus. These remain on the long tail.

For our current probabilistic ETH forecast, see Ethereum price prediction.

Coverage on The Daily Coins

FAQ

Is Ethereum still relevant in 2026? Ethereum remains the largest smart-contract platform by TVL, the dominant settlement layer for tokenised real-world assets, and the venue where most institutional capital chooses to settle when it comes on-chain. The relevance is institutional — Ethereum is not the fastest or cheapest chain, but it is the most credibly neutral and broadly trusted.

How does Ethereum staking work? ETH holders can stake their tokens to help secure the network and earn rewards. Solo staking requires 32 ETH and running consensus client software. Pooled staking via Lido, Rocket Pool, or other providers allows participation with any amount. Yields are currently 3-4% annually, with potential additional yield through restaking. See our Lido vs Rocket Pool comparison for details.

What is the difference between L1 and L2? L1 (Ethereum mainnet) is the base layer where consensus and final settlement happen. L2s (Arbitrum, Base, Optimism, etc.) are scaling solutions that process transactions off-chain and post compressed data back to L1 for security. L2 transactions are typically 100-1000x cheaper and much faster than L1 transactions, while still inheriting L1’s security.

Will ETH reach $10K? Quantitative forecasts produce wide ranges. The base case for the late 2020s involves ETH appreciation driven by continued tokenisation growth and the deflationary mechanics of high-activity periods, but the specific price level depends on many variables that no model can fully capture.

What is restaking and is it safe? Restaking allows staked ETH (directly or via LSTs) to be reused as cryptoeconomic security for additional services beyond Ethereum’s base consensus. EigenLayer is the largest restaking platform. The additional yield is attractive but the additional slashing risk is meaningful. Best practices are still being developed.

How is Ethereum different from Bitcoin? Bitcoin is a monetary network with a fixed supply, designed to store value over long horizons. Ethereum is a smart-contract platform designed to host applications. They answer different questions and are best held together rather than as substitutes. See our Bitcoin vs Ethereum comparison.

Is ETH a security? The SEC has historically declined to call ETH a security in its public statements. The treatment of staking yields and certain derivative products (LSTs, restaking tokens) has been more contested. The regulatory framework continues to evolve, and the practical effect for individual ETH holders has been clear-enough operational guidance.

What is account abstraction and why does it matter? Account abstraction allows externally-owned accounts (the typical user wallet) to behave more like smart-contract accounts — supporting batch transactions, gas sponsorship, social recovery, custom signature schemes, and other UX improvements that have historically been impossible without migrating to entirely new account types. EIP-7702, shipped in 2024, made this practically usable. The downstream effect has been a generation of wallets that hide most of the complexity of self-custody from users — Coinbase’s smart wallet, Privy, Dynamic, and others all leverage these primitives. For most new users coming on-chain in 2026, the experience is closer to a regular fintech app than to the seed-phrase-required experience of earlier years.

What is the difference between rollups, validiums, and volitions? Rollups post all transaction data to Ethereum mainnet, ensuring full data availability. Validiums use off-chain data availability (faster and cheaper but with weaker availability guarantees). Volitions let users choose per-transaction whether to use the rollup or validium data availability model. The choice involves a security trade-off — rollups have the strongest guarantees, validiums have weaker guarantees but lower costs. Most Ethereum L2s as of 2026 are rollups; some specialised chains use validium or volition models for cost reasons.

How do gas fees work in 2026? Base-layer gas fees on Ethereum mainnet vary with network demand from sub-dollar amounts during quiet periods to multiple dollars during congestion. L2 transactions typically cost a fraction of a cent. EIP-1559 — the fee market mechanism introduced in 2021 — uses a base fee that adjusts dynamically based on block fullness, plus a priority fee paid to validators. The base fee is burned, removing ETH from circulation. The combination of L2 adoption (reducing mainnet demand for typical transactions) and EIP-4844 (reducing L2 costs) has made the network meaningfully more affordable for users than at any point since 2020.

What is the relationship between ETH and the broader DeFi ecosystem? ETH is the base currency of the Ethereum DeFi ecosystem. It is the most-traded pair on every Ethereum-based DEX. It is the most-deposited collateral on lending protocols. It is the asset most-staked in liquid staking products. Many DeFi yield strategies are ETH-denominated rather than USD-denominated. This positioning means ETH benefits from growth in the broader DeFi ecosystem in ways that are not purely about transaction-fee revenue — ETH appreciates structurally as more economic activity is conducted in ETH-denominated terms.