Neura Intelligence Index

AI Model Benchmarks

BigBench-Hard — AI Model Leaderboard & Results | Neura Market

Back to Leaderboard

BigBench-Hard

Subset of BIG-Bench tasks that are challenging for language models — tests reasoning

reasoning0-100 percentPaper / Website

Model Rankings (12)

#	Model	Score	Provider	Source
	Claude Opus 4🇺🇸			aggregated
	o1🇺🇸			aggregated
	Claude Sonnet 4🇺🇸			aggregated
	Claude 3.5 Sonnet🇺🇸			aggregated
	GPT-4o🇺🇸			aggregated
	Claude 3 Opus🇺🇸			aggregated
	Llama 3.1 405B🇺🇸			aggregated
	GPT-4 Turbo🇺🇸			aggregated
	Qwen 2.5 72B🇨🇳			aggregated
	Llama 3.3 70B🇺🇸			aggregated
	Llama 3.1 70B🇺🇸			aggregated
	GPT-4o mini🇺🇸			aggregated