Terminal-Bench — cost vs performance

Best current row per canonical model in the group “accuracy · variant=v4.0 · evaluator=Artificial Analysis” (114 models) against context window (tokens). The dashed line is the Pareto frontier: no model is both better and cheaper than a point on it.

← Leaderboard

XOutput price Input price Parameters Context Memory (est.)Scalelog linearBubblecontext params none

open / restricted weights closed Pareto frontier (9 models)bubble = context window

Frontier models 9

Pareto frontier
Model	accuracy on Terminal-Bench	Rank	context window (tokens)	Provider	Trust
gpt-6-astraOpenAI · Closed	59.6%	1	1.05M	—	Independent
Claude Fable 5.1Anthropic · Closed	55.0%	2	1M	—	Independent
Grok 4.6xAI · Closed	21.2%	13	500K	—	Independent
gpt-5-5-instant-06-26OpenAI · Closed	12.6%	21	400K	—	Independent
GLM 5.1Z.ai (Zhipu AI) · Open weights	2.02%	39	204.8K	—	Independent
Muse Glimmer 30BMeta AI · Open weights	0.51%	49	131.1K	—	Independent
qwen3-14b-instructAlibaba Group · Open weights	0%	60	32.8K	—	Independent
qwen3-32b-instructAlibaba Group · Open weights	0%	60	32.8K	—	Independent
qwen3-8b-instructAlibaba Group · Open weights	0%	60	32.8K	—	Independent

Methodology. Points are the best current row per canonical model in comparability group 'accuracy · variant=v4.0 · evaluator=Artificial Analysis'. Price = cheapest current offer across providers (the provider shown). Pareto frontier maximises the score and minimises x; exact ties are all kept. memory_estimate is an estimate (see /methodology). Only the selected comparability group is plotted; points under other configurations are not mixed in. Price = cheapest current offer across providers at the time of the last crawl. Nothing is estimated except where marked.