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

OfficePrism-420 (arXiv 2611.22262) single-domain enterprise agent evaluation (ReAct (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 K376.2#62
2Claude Opus 4.774#86
3Claude Opus 4.874#64
4Claude Sonnet 573.1#124
5DeepSeek V4 Pro72.1#201
6GLM-5.271.9#154
7GLM-5.171.4#200
8GPT-5.571#56
9DeepSeek V4 Flash71#273
10Qwen 3.7 Plus70#221
11Qwen 3.6 Max65#192
12Qwen 3.6 Plus63.8#245
13Kimi K2.663.3#197
14Qwen 3.7 Max63.3#160

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-react-search-nl-db-pass · How It Works · Data refreshed daily, snapshot 2026-11-22.