τ²-bench — cost vs performance

Best current row per canonical model in the group “pass^1 · variant=Telecom · evaluator=Artificial Analysis” (308 models) against ESTIMATED memory at 4-bit, 8K context (GB). 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 (7 models)bubble = context windowEstimated

Frontier models 7

Pareto frontier
Model	pass^1 on τ²-bench	Rank	ESTIMATED memory at 4-bit, 8K context (GB)	Provider	Trust
Z.ai GLM 5.2Z.ai (Zhipu AI) · Open weights	99.1%	1	433.7 GBest.	—	Independent
GLM 4.7 FlashZ.ai (Zhipu AI) · Open weights	98.8%	3	18.5 GBest.	—	Independent
Qwen3.6 27BQwen · Open weights	94.2%	29	16.5 GBest.	—	Independent
Qwen3.5-4BQwen · Open weights	92.1%	45	3.2 GBest.	—	Independent
Qwen3-VL-4B-InstructQwen · Open weights	23.4%	218	3.1 GBest.	—	Independent
Llama-3.2-3BMeta AI · Restricted weights	21.1%	232	2.4 GBest.	—	Independent
LFM2-1.2BLiquid AI · Open weights	12.6%	267	1.2 GBest.	—	Independent

Methodology. Points are the best current row per canonical model in comparability group 'pass^1 · variant=Telecom · 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.