SGR-Bench - Constraint-Guided: leaderboard
Metric: Item-level F1 (%) on the constraint-guided formulation (the retrieval constraints are stated explicitly), 100 state-gated retrieval tasks on specialized data sites (scholarly, official statistics, life science, environmental, regulatory); eight models in production CLI agents (GPT-5.5 in Codex CLI, the others in Claude Code) with the same Serper search, fetch and PDF tools, plus three commercial search agents; outputs parsed into rows and aligned by key fields; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 68.22 |
| 2 | Claude Opus 4.7 (Medium) | 64.35 |
| 3 | GLM-5.1 | 63.17 |
| 4 | Deep Research | 57.27 |
| 5 | DeepSeek V4 Pro | 56.67 |
| 6 | Gemini 3.1 Pro (Preview) (Medium) | 56.43 |
| 7 | Kimi K2.5 (Thinking) | 47.54 |
| 8 | Qwen 3.6 Plus | 44.25 |
| 9 | Seed 2.0 Pro (Medium) | 31.48 |
Interactive version: theaggregate.ai/benchmark?slug=sgr-bench-constraint-guided · How It Works · Data refreshed daily, snapshot 2026-10-07.