Verkle Tree

A Verkle tree gets its name from combining "vector commitment" with "tree," and it exists to solve a specific problem: as Ethereum's blockchain state grows into hundreds of gigabytes, proving that a single account balance or contract slot is correct becomes increasingly expensive. A traditional Merkle tree, or the Merkle Patricia trie Ethereum uses today, requires submitting every sibling hash along the path from leaf to root, so proofs grow larger as the tree grows. Verkle trees replace that chain of hashes with a single polynomial-based vector commitment per node, letting one compact proof attest to many pieces of data at once.

Each node commits to all of its children using cryptography such as KZG polynomial commitments, so a witness proving a leaf's inclusion needs only one opening per level instead of a full list of sibling hashes. Because Ethereum's tree has a wide branching factor, this can shrink proofs that would otherwise run to several kilobytes down to a few hundred bytes, even against a state holding billions of entries.

This matters for scalability because full nodes today must store the entire state locally just to validate blocks. With Verkle trees, a stateless client could instead receive a block bundled with a small cryptographic witness, check it against the tree's root hash, and validate the block without ever holding gigabytes of account data. That lowers the hardware bar for running a node or solo-staking validator, helping keep the network decentralized as usage grows.

Verkle trees were designed as the centerpiece of Ethereum's Verge upgrade phase, but the plan has since become less certain: researchers have proposed binary hash trees paired with SNARK proofs as an alternative, since that approach avoids the elliptic-curve cryptography that would leave Verkle commitments exposed to future quantum computers. As of mid-2026, Ethereum's core developers are still evaluating both designs on testnets, and no final choice for replacing the current state trie has been locked in.