CLE - Hermes-Agent Trial 1 Level 4: leaderboard

Full Continual Learning Exam: one serial stream of 120 professional tasks, five levels each (600 levels), in fixed order. Trial 1 permits one submission per level; Trial 2 permits one feedback-guided revision. Unreached levels fail. Lk is cumulative clearance through level k. DeepSeek V4 Flash grades the human-authored rubrics. Harness/provider default effort is unspecified. These are the full exam results, not the 20-task CLE Lite release. Harness: Hermes-Agent; trial: t1. Metric: Cumulative pass rate (%).

Metric: Cumulative pass rate (%). Source: haoyu-zhao.github.io. Saturation forecast: Around 2030. 6 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 3.7 Flash10.83#22
2GLM-5.37.5#39
3Hy36.67#96
4DeepSeek V4 Pro5.83#109
5Qwen 3.8 Max5.83#32

No result here: #5 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=cle-hermes-agent-trial-1-level-4 · How It Works · Data refreshed daily, snapshot 2026-10-26.