WildClawBench

mentions 1 type Organization feed RSS

// recent coverage 1 mentions

04:50

2026-08-17

runtimewire.com

artificial-intelligence

WildClawBench ranks Alibaba's Qwen3.8-27B 15th on agent tasks

Alibaba's Qwen3.8-27B ranked 15th among 33 models in WildClawBench, scoring 48.0% across 60 agent tasks run through the OpenClaw harness, according to the third-party benchmark. The 27-billion-paramet…

// co-occurs with top 7 entities

Alibaba 1 Qwen3.8-27B 1 OpenClaw 1 OpenAI 1 Anthropic 1 xAI 1 Moonshot AI 1

// topics top 3 topics

artificial intelligence 1 ai agents 1 large language models 1