OfficePrism-420 - DeepAgents (Search) - NL-DB Pass: leaderboard

OfficePrism-420 (arXiv 2611.22262) single-domain enterprise agent evaluation (DeepAgents (Search) - NL-DB Pass) across 420 tasks in 7 business functions.

Metric: Joint NL-DB Pass Rate (%) over 420 single-domain enterprise tasks. Source: arxiv.org. Saturation forecast: Around February 2027. 14 models tracked. Updated 22 Nov 2026.

Top models

#ModelScoreOverall rank
1Kimi K376#62
2Claude Opus 4.772.6#86
3Claude Sonnet 572.6#124
4GPT-5.571.9#56
5Claude Opus 4.871.2#64
6Qwen 3.7 Plus71#221
7GLM-5.270.7#154
8Qwen 3.7 Max70.2#160
9GLM-5.169#200
10DeepSeek V4 Pro68.1#201
11Qwen 3.6 Max67.4#192
12Kimi K2.662.9#197
13DeepSeek V4 Flash62.9#273
14Qwen 3.6 Plus58.3#245

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=officeprism-420-deepagents-search-nl-db-pass · How It Works · Data refreshed daily, snapshot 2026-11-22.