Ethereum, Solana, Base, and Avalanche upgrades in 2026
In the second half of 2026, several major blockchain networks are preparing or implementing significant technical upgrades. Ethereum, Solana, Avalanche, and Base are preparing some of the biggest upgrades in recent years. Tim Sun, a senior researcher at Hong Kong-based HashKey Group, explained that protocol upgrades so far have mainly focused on new features, speed, and bandwidth. He says that the focus in 2026 is shifting towards reliability, predictable management, and infrastructure that can support large-scale financial applications at the institutional level.
Source: cointelegraph
Ethereum: Glamsterdam
Glamsterdam is one of Ethereum’s more significant upgrades planned for this year and is already being tested on the Platåberget public testnet, launched on August 17, 2026. The goal is to improve scalability, strengthen the first layer of the network, and simplify use, with a mainnet launch expected in the last quarter of 2026. The upgrade includes two key proposals. The built-in separation of block proposers and builders, known as ePBS, should reduce the concentration of control over the order of transactions, which today depends on a small number of specialized builders. Another proposal, Block Level Access Lists, allows the network to process transactions in parallel. According to Holly Atkinson of 1inch, it is the largest Ethereum upgrade since the transition to proof of stake in September 2022. Pavan Kaur, a judge at the Solana Foundation and founder of RuleSpark, warns that ePBS does not completely solve the problem of MEV, i.e. the additional earnings that validators extract from the order of transactions, or the centralization of block builders. According to her, practices such as so-called sandwich attacks are more likely to move than disappear.
Source: cointelegraph
Solana: Alpenglow
Alpenglow is the biggest change for Solana this year, a thorough overhaul of the network’s consensus mechanism. It was approved by the community in September 2025, and according to the latest plans, it should be activated in October 2026 through the Agave 4.3 client. As of mid-August 2026, this client is still in the alpha phase and officially marked as unsuitable for production. Instead of the current TowerBFT mechanism, Alpenglow is introducing a new system built around the Votor and Rotor components. The goal is to reduce the time required for transaction finality from approximately 12.8 seconds to about 100 to 150 milliseconds. The upgrade also eliminates on-chain validator voting transactions, which today make up a large part of network traffic. Hadley Stern, a member of the board of DeFi Development Corp, believes that the removal of these voting transactions is real news for institutional investors, as it refines the validator economy and gives a more accurate picture of the state of the network, which, in his opinion, is also relevant for institutions that analyze the use of SOL as part of the corporate treasury.
Source: cointelegraph
Base: Beryl
Base activated its Beryl hard fork at the end of June 2026. The upgrade introduces B20, a new token standard, shortens the deadline for payment finality from seven to five days, and integrates Reth V2, with the aim of reducing data storage requirements on the node and speeding up transaction execution.
The activation came shortly after a brief block production shutdown, which, according to Base’s subsequent root cause analysis, was due to a sequencer error. Base co-founder Jesse Pollak confirmed that customer funds were not affected, but added that network downtime is never acceptable and that he will use the lessons learned from the incident to strengthen Base as a 0-to-24 finance platform.
Source: cointelegraph
Avalanche: Focus on Institutions
Avalanche has already implemented the biggest structural change, the Etna hard fork, by replacing the old model of subnets with sovereign Avalanche L1 chains. According to Tim Sun, this has reduced the cost of running its own blockchain by more than 99 percent, reducing the cost of running its own blockchain networks on Avalanche infrastructure.
One example is Progmat, a platform that, according to Sun, holds about 63 percent of the Japanese tokenized securities market and has transferred more than $2 billion in tokenized assets to its own Avalanche L1. Along with this is the Avalanche Payments Collective, which includes companies such as Franklin Templeton, VanEck and WisdomTree.
Avalanche is also working on two upgrades aimed at increasing the performance of its C-Chain, i.e. environments compatible with Ethereum’s virtual machine. Streaming Asynchronous Execution separates transaction execution from consensus, with the goal of increasing throughput and reducing fee fluctuations during periods of increased network activity.
Source: cointelegraph
Bitcoin: the biggest question mark
Bitcoin is an exception in this society. Its biggest topics in 2026 are not scheduled upgrades, but the continuation of long discussions about whether the network should become more programmable and how urgently it should be protected from quantum computers. The last major soft fork, Taproot, was activated in November 2021.
Since then, proposals such as OP_CAT, CheckTemplateVerify, and the Lightning solution LNHANCE have been discussed, but none of them has an agreed path to activation. The situation is similar with BIP-360, a proposal to move to quantum-resistant addresses, whose authors estimate that a full migration would take years and in the best-case scenario. An activation deadline of the end of 2026 has not currently been agreed for these proposals.
While Ethereum, Solana, Base and Avalanche have specific plans for individual upgrades, with Bitcoin, these proposals are still in various stages of technical discussion and development.