Benchmarks / Agentic

EdgeBench

Uses the leaderboard's overall score at the longest (12-hour) environment-learning horizon. Score: 12-hour-horizon overall score. Collected by Jake Boggs' benchmark dashboard.

5models

–Open Index weight

–last updated

boggspublisher

full table below

#	Model	Effort	Score	Cost / task	Harness
1	Opus 4.8	max	51.3%	–	–
2	GPT-5.5	xhigh	48.4%	–	–
3	GPT-5.4	xhigh	39.3%	–	–
4	GLM 5.1	max	37.4%	–	–
5	DeepSeek V4 Pro	max	31.0%	–	–