Neura Intelligence Index

AI Model Benchmarks

Back to Leaderboard

TerminalBench

Terminal/CLI task completion — executing shell commands to achieve goals.

coding0-100 percent

Model Rankings (25)

#	Model	Score	Provider	Source
	Claude Sonnet 4.5🇺🇸			verified
	MiniMax M2.1🇨🇳			verified
	M Kimi K2-Thinking-0905🇨🇳			verified
	MiniMax M2🇨🇳			verified
	Claude Opus 4.1🇺🇸			verified
	Nova 2 Pro🇺🇸			verified
	Claude Haiku 4.5🇺🇸			verified
	Z GLM-4.6🇨🇳			verified
	M LongCat-Flash-Chat🇨🇳			verified
	Claude Opus 4🇺🇸			verified
	DeepSeek-V3.2-Exp🇨🇳			verified
	Z GLM-4.5🇨🇳			verified
	Claude Sonnet 4🇺🇸			verified
	Claude 3.7 Sonnet🇺🇸			verified
	M LongCat-Flash-Lite🇨🇳			verified
	Z GLM-4.7🇨🇳			verified
	Nova 2 Lite🇺🇸			verified
	DeepSeek-V3.1🇨🇳			verified
	X MiMo-V2-Flash🇨🇳			verified
	Z GLM-4.5-Air🇨🇳			verified
	M Kimi K2 Instruct🇨🇳			verified
	Nemotron 3 Super (120B A12B)🇺🇸			verified
	M Kimi K2-Instruct-0905🇨🇳			verified
	Nemotron 3 Nano (30B A3B)🇺🇸			verified
	DeepSeek-R1-0528🇨🇳			verified

TerminalBench — AI Model Leaderboard & Results | Neura Market