Horizon - Medium — leaderboard

Metric: Task Pass Rate (%). Source: www.orinlabs.ai. 24 models tracked.

Top models

#ModelScore
1OpenClaw (LCM) - claude-opus-4.870.77
2OpenClaw (LCM) - gpt-5.561.54
3OpenClaw (LCM) - gemini-3.5-flash60
4RLM - claude-opus-4.858.46
5OpenClaw (LCM) - gemini-3.1-pro-preview55.38
6RLM - gpt-553.85
7RLM - claude-sonnet-4.653.85
8Codex - gpt-5-codex46.15
9OpenClaw (LCM) - deepseek-v4-pro43.08
10Claude Code - claude-opus-4.841.54

Interactive version: theaggregate.ai/benchmark?slug=horizon-medium · How the rankings work · Data refreshed daily, snapshot 2026-07-22.