SGR-Bench - Constraint-Guided: leaderboard

Item-level F1 in percent for search agents on 100 state-gated retrieval tasks, where answer-bearing records on specialized data sites appear only after the right filters and selections are set (SGR-Bench, 2026), in the formulation that states the retrieval constraints explicitly. Higher is better.

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 February 2027. 11 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1GPT-5.568.22#54
2Claude Opus 4.7Medium64.35#85
3GLM-5.163.17#198
4Deep Research57.27
5DeepSeek V4 Pro56.67#196
6Gemini 3.1 Pro (Preview)Medium56.43#99
7Kimi K2.5Thinking47.54#266
8Qwen 3.6 Plus44.25#238
9Seed 2.0 ProMedium31.48#222
10Gemini Deep Research29.72
11Google Search AI Mode16.73

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=sgr-bench-constraint-guided · How It Works · Data refreshed daily, snapshot 2026-11-21.