Horizon - RAG - Medium: leaderboard
Metric: Task Pass Rate (%). Source: www.orinlabs.ai. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | gpt-5.5 | 36.92 |
| 2 | claude-opus-4.8 | 33.85 |
| 3 | claude-sonnet-4.5 | 30.77 |
| 4 | claude-haiku-4.5 | 23.08 |
| 5 | gpt-5-mini | 12.31 |
| 6 | gemini-3.5-flash | 6.15 |
Interactive version: theaggregate.ai/benchmark?slug=horizon-rag-medium · How It Works · Data refreshed daily, snapshot 2026-10-09.