ProfBench (Search): leaderboard
Professional work task benchmark search subset reported in provider launch comparisons for agentic knowledge-work performance.
Metric: Score (%). Source: developer.nvidia.com. Source: benchmarklist.com. Saturation forecast: Around August 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Pro | 59.9 | #102 |
| 2 | DeepSeek V4 Flash | 57 | #137 |
| 3 | Kimi K2.6 | 56 | #104 |
| 4 | Nemotron 3 Ultra | 56 | #235 |
| 5 | MiniMax-M2.7 | 52 | #258 |
| 6 | GLM-5.1 | 46 | #97 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=profbench-search · How It Works · Data refreshed daily, snapshot 2026-10-20.