KAware - External Function: leaderboard

The knowing-acting score of KAware (2026) on tasks that can only be solved with external tools: the harmonic mean of how well the agent states the needed tools and how well its actual tool calls match them, in percent; higher is better. KAware (2026) keeps tools available on every task, so an agent must judge whether it needs them.

Metric: The 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), on the external function setting, 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: Estimated already saturated. 18 models tracked.

Top models

#ModelScoreOverall rank
1Qwen 3 235B A22B 2507 (Thinking)97.42#463
2Qwen 3 Max (Preview)97.36#447
3Qwen 3.5 397B A17B97.34#246
4Claude Sonnet 4.5 (Thinking)97.3#250 (Claude Sonnet 4.5)
5Qwen 3.5 397B A17B (Non-reasoning)96.62#246 (Qwen 3.5 397B A17B)
6DeepSeek V4 Flash (Non-reasoning)96#258 (DeepSeek V4 Flash)
7GPT-5.595.9#49
8Claude Sonnet 4.595.62#250
9Gemini 3 Pro (Preview)91.75#113
10GPT-4.191.26#495
11GPT-4o (2024-11-20)90.98#866
12DeepSeek V3.2 (Non-reasoning)90.84#406 (DeepSeek V3.2)
13Gemini 3 Flash (Preview)89.61#144
14O4 Mini (2025-04-16)89.22#365
15GPT-588.36#167

No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.

Interactive version: theaggregate.ai/benchmark?slug=kaware-external-function · How It Works · Data refreshed daily, snapshot 2026-11-13.