// BENCHMARK

EvalPlus benchmark

AI model leaderboard for the EvalPlus benchmark. Compare how large language models score on EvalPlus, see the full ranking, and understand what this AI benchmark measures. Kimi Linear currently leads with 61. EvalPlus harness — augments HumanEval/MBPP with adversarial test cases that expose surface-level passes.

Leaderboard

# Model Organization Score Variant Source
#1 Kimi Linear Moonshot 61 sft official ↗
#2 Qwen2-57B-A14B Qwen 57.2 base-from-hf-readme official ↗
#3 Qwen2-7B Qwen 54.2 base-from-hf-readme official ↗
#4 Yi-1.5-34B 01.AI 51.9 cited-qwen2-57b official ↗
#5 Qwen1.5-32B Qwen 50.4 cited-qwen2-57b official ↗
#6 Mixtral 8x7B Mistral AI 46.4 cited-qwen2-57b official ↗

Frequently asked questions about EvalPlus

What is the EvalPlus benchmark?

EvalPlus harness — augments HumanEval/MBPP with adversarial test cases that expose surface-level passes.

How is the EvalPlus benchmark scored?

EvalPlus is scored using the pass@1 metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.

Which AI model scores highest on EvalPlus?

As of the latest reported scores on GenAIList, Kimi Linear achieves the highest result on EvalPlus with a score of 61.

Is a higher EvalPlus score better?

Yes. On EvalPlus a higher score indicates better performance, so models near the top of the leaderboard are the strongest.