Prototype

Work published while it is still being tested and refined. These are ideas in motion, not finished conclusions.

ECDSA.fail

Finding smaller quantum circuits for attacking elliptic curve signatures. Solvers surpassed Google Quantum AI's benchmark within hours of launch.

Improving verifiable solutions to hard, unsolved problems in computer science. Open to any solver willing to push the frontier forward.

Pushing open-model inference past its current baseline on Apple Silicon. A joint challenge with Poolside AI to optimize the Laguna model.

Speeding up Flock, Succinct's post-quantum proof system for Ethereum. Progress here compounds directly into faster, cheaper zero-knowledge proofs.

Raising the throughput of the Lighter Exchange. Solvers compete to move real infrastructure, not just benchmarks.

Making MODEXP and RIPEMD-160 cheaper onchain without weakening their proofs. Every gain lowers cost for every transaction that depends on them.