SWE-bench Verified — cost vs performance

Best current row per canonical model in the group “resolved · board=Verified · system=mini-SWE-agent” (42 models) against cheapest current input price (USD / 1M 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

Group

open / restricted weights closed Pareto frontier (6 models)bubble = context window

Frontier models 6

Pareto frontier
Model	resolved on SWE-bench Verified	Rank	cheapest current input price (USD / 1M tokens)	Provider	Trust
Claude Opus 4.5Anthropic · Closed	76.8%	1	$2.5	Anthropic API	Official board
MiniMax M2.5MiniMax · Open weights	75.8%	2	$0.30	MiniMax API	Official board
DeepSeek V3.2DeepSeek · Open weights	70%	11	$0.269	OpenRouter	Official board
MiniMax M2MiniMax · Open weights	61%	20	$0.255	MiniMax API	Official board
gpt-5-miniOpenAI · Closed	56.2%	24	$0.125	OpenAI API	Official board
gpt-5-nanoOpenAI · Closed	34.8%	34	$0.025	OpenRouter	Official board

Methodology. Points are the best current row per canonical model in comparability group 'resolved · board=Verified · system=mini-SWE-agent'. 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.