LLM Stats (Seal-0): leaderboard
Search-oriented benchmark for evaluating agentic information retrieval, tool use, and answer synthesis over hard search tasks. Scores are reported by LLM Stats.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around January 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2.5 | 57.4 | #144 |
| 2 | Qwen 3.5 27B | 47.2 | #187 |
| 3 | Qwen 3.5 397B A17B | 46.9 | #146 |
| 4 | Qwen 3 Max (Thinking) | 46.9 | #208 (Qwen 3 Max) |
| 5 | Qwen 3.5 122B A10B | 44.1 | #173 |
| 6 | Qwen 3.5 35B A3B | 41.4 | #256 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-seal-0 · How It Works · Data refreshed daily, snapshot 2026-10-20.