SlopCodeBench

Uses the published best-version leaderboard and Iso Solve score. Score: isolated checkpoints solved. Collected by Jake Boggs' benchmark dashboard.

13models

–Open Index weight

–last updated

boggspublisher

#	Model	Effort	Score	Cost / task	Harness
1	GPT-5.5	xhigh	28.1%	–	–
2	GPT-5.3 Codex	xhigh	26.0%	–	–
3	GPT-5.4	xhigh	23.5%	–	–
4	GPT-5.2 Codex	xhigh	21.9%	–	–
5	Opus 4.6	max	20.9%	–	–
6	Opus 4.7	max	20.9%	–	–
7	Kimi K2.6	max	18.9%	–	–
8	Opus 4.5	high	17.4%	–	–
9	Sonnet 4.6	max	16.8%	–	–
10	GLM 5.1	max	13.8%	–	–
11	GPT-5.4 Mini	xhigh	13.8%	–	–
12	Kimi K2.5	max	9.7%	–	–
13	MiniMax M2.7	high	4.1%	–	–