Trial by Trace - OpenClaw - Agent Pass: leaderboard

Trial by Trace measures model task performance under a named native agent harness on 95 mechanism-focused tasks per model. Agent Pass requires all agent task criteria to pass; Agent Score is criterion-level performance. Harness-only scores, costs and mixed Agent/Harness composites are not admitted. arXiv v1; 95 tasks per model/harness; one selected execution including recovery replacements; DeepSeek-V4-Flash semantic judge with variable scoring budgets; native harness versions, model API snapshots and reasoning effort unreported. Metric: Agent Pass (%), higher is better.

Metric: Agent Pass (%). Source: tracelite.ai. Saturation forecast: Rough model projection: around 2026. 9 models tracked.

Top models

#ModelScoreOverall rank
1GPT-5.6 Sol (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)42.1
2GLM-5.3 (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)41.1
3Claude Opus 5 (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)41.1
4Kimi K3 (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)40
5DeepSeek V4 Pro (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)38.9
6Gemini 3.7 Flash (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)38.9
7Qwen 3.8 Max (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)38.9
8Grok 4.6 (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)33.7
9Hy3 (Benchmark Sweep Run: source=Trial-by-Trace; harness=OpenClaw; snapshot-and-effort=unreported)29.5

Interactive version: theaggregate.ai/benchmark?slug=trial-by-trace-openclaw-agent-pass · How It Works · Data refreshed daily, snapshot 2026-11-06.