Neura Intelligence Index

AI Model Benchmarks

HumanEval — AI Model Leaderboard & Results | Neura Market

Back to Leaderboard

HumanEval

Code generation benchmark — 164 Python programming problems measuring functional correctness

coding0-100 pass_rateSaturatedPaper / Website

Model Rankings (19)

#	Model	Score	Provider	Source
	Claude Opus 4🇺🇸			aggregated
	Claude Sonnet 4🇺🇸			aggregated
	o1🇺🇸			aggregated
	Claude 3.5 Sonnet🇺🇸			aggregated
	Mistral Large 2🇫🇷			aggregated
	GPT-4o🇺🇸			aggregated
	Llama 3.1 405B🇺🇸			aggregated
	Llama 3.3 70B🇺🇸			aggregated
	Claude 3.5 Haiku🇺🇸			aggregated
	GPT-4 Turbo🇺🇸			aggregated
	GPT-4o mini🇺🇸			aggregated
	Qwen 2.5 72B🇨🇳			aggregated
	Claude 3 Opus🇺🇸			aggregated
	Llama 3.1 70B🇺🇸			aggregated
	Mixtral 8x22B🇫🇷			aggregated
	Nemotron 4 340B🇺🇸			aggregated
	Command R+🇨🇦			aggregated
	Llama 3.1 8B🇺🇸			aggregated
	Phi-3 Medium🇺🇸			aggregated