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

OfficePrism-420 (arXiv 2611.22262) single-domain enterprise agent evaluation (DeepAgents (Full) - 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 Sonnet 571.2#124
3GLM-5.270.2#154
4Claude Opus 4.770#86
5Claude Opus 4.870#64
6GPT-5.568.1#56
7Qwen 3.6 Plus67.9#245
8Qwen 3.7 Max67.9#160
9Qwen 3.7 Plus67.9#221
10GLM-5.167.1#200
11DeepSeek V4 Pro66.2#201
12DeepSeek V4 Flash66.2#273
13Kimi K2.665#197
14Qwen 3.6 Max63.6#192

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