KAware: leaderboard

Metric: Average knowing-acting score (%, KAS): harmonic mean of knowing accuracy (Jaccard overlap between the tools the agent says the task needs and the gold tool set) and acting accuracy (Jaccard overlap between the tools it actually calls and the gold set), task-weighted over the three settings, on the KAware tool-use tasks, where tools are always available but the task may need them (external function, 310 tasks), mix them with parametric knowledge (hybrid composition, 368 tasks) or be solvable from knowledge alone (internal function, 398 tasks); higher is better. Source: arxiv.org. Saturation forecast: Around June 2027. 18 models tracked.

Top models

#ModelScore
1GPT-569.73
2GPT-5.565.55
3O4 Mini (2025-04-16)64.15
4Gemini 3 Pro (Preview)62.9
5Qwen 3.5 397B A17B56.92
6Qwen 3.5 397B A17B (Non-reasoning)55.72
7GPT-4.154.75
8Gemini 3 Flash (Preview)53.61
9DeepSeek V4 Flash (Non-reasoning)53.4
10Claude Sonnet 4.553.34
11GPT-4o (2024-11-20)53.1
12Qwen 3 Max (Preview)51.99
13DeepSeek R1 052849.8
14DeepSeek V3.2 (Non-reasoning)49.14
15Qwen 3 235B A22B 2507 Instruct42.43

Interactive version: theaggregate.ai/benchmark?slug=kaware · How It Works · Data refreshed daily, snapshot 2026-09-29.