SciCode — cost vs performance

Best current row per canonical model in the group “accuracy · evaluator=Artificial Analysis” (129 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 (8 models)bubble = context windowEstimated

Frontier models 8

Pareto frontier
Model	accuracy on SciCode	Rank	ESTIMATED memory at 4-bit, 8K context (GB)	Provider	Trust
Kimi K3Moonshot AI · Open weights	59.5%	5	1.6K GBest.	—	Independent
GLM 5.3Z.ai (Zhipu AI) · Open weights	59.0%	6	433.7 GBest.	—	Independent
GLM 5.3 FlashZ.ai (Zhipu AI) · Open weights	51.6%	28	185.3 GBest.	—	Independent
MiniMax M2.7MiniMax · Open weights	50.1%	37	132 GBest.	—	Independent
Qwen3.8 27BQwen · Open weights	46.6%	51	16.5 GBest.	—	Independent
gpt-oss-20bOpenAI · Open weights	38.9%	86	12.5 GBest.	—	Independent
Qwen3 8BQwen · Open weights	27.9%	115	5.2 GBest.	—	Independent
LFM2.5-2.6B (free)Liquid AI · Open weights	14.3%	129	2.1 GBest.	—	Independent

Methodology. Points are the best current row per canonical model in comparability group 'accuracy · 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.