a16z Crypto Unveils Quantum-Resistant zkVM Lattice Jolt

a16z Crypto Unveils Quantum-Resistant zkVM Lattice Jolt

By: WEEX|2026/09/10 00:49:48

WEEX View

  1. The main variable to watch is whether developers treat Lattice Jolt as a research milestone or begin testing it in production-oriented zk workflows. Technical performance claims matter more if they translate into broader tooling and integration.
  2. Its design choice is also notable: replacing an elliptic curve-based commitment method with a lattice-based one shifts attention toward post-quantum security in zk infrastructure, an area that has drawn interest but limited mainstream deployment.
  3. Near-term follow-up should center on independent benchmarking, compatibility with existing zk stacks, and progress on the roadmap for added zero-knowledge features and a hash-based Jolt version.

Andreessen Horowitz Crypto has unveiled Lattice Jolt, an open-source zero-knowledge virtual machine that uses quantum-resistant cryptography and, according to the firm, boosts prover and verifier speed by about 2-3 times while cutting proof size to under 100KB.

Lattice Jolt is designed to prove that a program was executed in a predetermined way, allowing developers to verify the correctness of execution results. Andreessen Horowitz Crypto said the system targets 128-bit quantum-resistant security, positioning it as a post-quantum alternative within zero-knowledge infrastructure.

The firm said the new zkVM replaces Dori, its earlier elliptic curve-based polynomial commitment method, with Akita, a lattice-based commitment scheme built on the modular SIS assumption. According to the announcement, Akita supports a transparent setup and is intended to provide quantum-resistant security.

Andreessen Horowitz Crypto also disclosed a set of performance metrics for the release. It said the CPU version of Lattice Jolt processes more than 2 million RISC-V cycles per second, while Apple Metal-based GPU acceleration pushes throughput above 10 million cycles per second. The firm added that prover memory use fell from about 300 bytes per cycle to 200 bytes.

The release is being positioned as open-source infrastructure rather than a token or network launch. Andreessen Horowitz Crypto said it plans to add zero-knowledge features in future updates and is also working on a hash-based version of Jolt. The announcement did not specify deployment timelines, outside adopters, or production integrations.

Why It Matters

The release highlights a more specialized direction in crypto infrastructure: improving zero-knowledge performance without ignoring long-horizon security assumptions. For teams building verifiable compute systems, zk tooling, or rollup-related infrastructure, the combination of smaller proofs, lower memory use, and post-quantum design could make this line of research more relevant.

It also puts more attention on whether quantum-resistant cryptography can move from theory-heavy research into practical developer tools. If projects begin testing lattice-based components in real zk environments, that could widen the design space for proof systems beyond current elliptic curve-based approaches.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.

About WEEX View

WEEX View is a crypto analysis and intelligence hub, covering the latest in Web3, AI, and global markets. Get independent research and in-depth insights to stay ahead of market trends and trading opportunities.

-- Price

--
--
--
iconiconiconiconiconiconicon
Customer Support:@weikecs
Business Cooperation:@weikecs
Quant Trading & MM:bd@weex.com
VIP Program:support@weex.com