AlgoVeri: An Aligned Benchmark for Verified Code Generation on Classical Algorithms

Introduction

We introduce AlgoVeri, a benchmark that evaluates vericoding of $77$ classical algorithms in Dafny, Verus, and Lean. By enforcing identical functional contracts, AlgoVeri reveals critical capability gaps in current models. While frontier models achieve tractable success in Dafny ($40.3$% for Gemini-3 Flash), where high-level abstractions and SMT automation simplify the workflow, performance collapses under the systems-level memory constraints of Verus ($24.7$%) and the explicit proof construction required by Lean (7.8%). Beyond aggregate metrics, we uncover a sharp divergence in test-time compute dynamics: Gemini-3 effectively utilizes iterative repair to boost performance (e.g., tripling pass rates in Dafny), whereas GPT-OSS saturates early. Finally, our error analysis shows that language design affects the refinement trajectory: while Dafny allows models to focus on logical correctness, Verus and Lean trap models in persistent syntactic and semantic barriers.

Quick Start

All of the instructions for environment setup and the evaluation pipeline is released at https://github.com/haoyuzhao123/algoveri

AlgoVeri Benchmark

We provide three splits: dafny, verus, lean. Each split consist of 77 problems.

Citation

@article{zhao2026algoveri,
  title={AlgoVeri: An Aligned Benchmark for Verified Code Generation on Classical Algorithms},
  author={Zhao, Haoyu and Yang, Ziran and Li, Jiawei and He, Deyuan and Li, Zenan and Jin, Chi and Veeravalli, Venugopal V and Gupta, Aarti and Arora, Sanjeev},
  journal={arXiv preprint arXiv:2602.09464},
  year={2026}
}