KWBench - Mean Score: leaderboard
Metric: Mean task score (%): gated rubric score (0 when a mandatory criterion fails, otherwise 0.4 plus credit for good-to-have and ideal criteria) averaged over all evaluated tasks, over KWBench's 223 practitioner-sourced knowledge-work tasks (acquisitions, negotiations, clinical pharmacy, organizational politics, fraud analysis, incentive design), each encoding a game-theoretic pattern the model must recognize unprompted from raw inputs, with a code interpreter; a Gemini 3 Flash judge checks 15 binary rubric criteria per task, any failed mandatory criterion scores the task zero; best of three runs per model at provider-recommended temperature; higher is better. Source: arxiv.org. Saturation forecast: Around 2031. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 22.6 |
| 2 | GPT-5.4 | 17.7 |
| 3 | GLM-5-Turbo | 16.8 |
| 4 | Qwen 3.5 Plus | 12.1 |
| 5 | Gemini 3.1 Pro (Preview) | 12 |
| 6 | Kimi K2.5 | 11.3 |
| 7 | MiniMax-M2.7 | 11.1 |
| 8 | Gemini 3 Flash | 10.7 |
| 9 | GPT-5.4 (xHigh) | 9.7 |
| 10 | MiniMax-M2.5 | 8.5 |
| 11 | GPT-OSS-120B | 8.1 |
Interactive version: theaggregate.ai/benchmark?slug=kwbench-mean-score · How It Works · Data refreshed daily, snapshot 2026-10-07.