WildClawBench

What is WildClawBench?

WildClawBench evaluates AI agents across six task categories: productivity flow, code intelligence, social interaction, search & retrieval, creative synthesis…

Released

2026-04-07

Evaluates

智能体, Agent, 代码, Code, 创作, Creation, Agent Evaluation, In-the-Wild Tasks, Multimodal & Tool-Use Reasoning, 代码工程, 语言生成, Generation, 任务执行, Task Execution, 跨模态推理, Cross-modal Reasoning, 不支持, Unsupported

Openness

open

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/WildClawBench

WildClawBench code and dataset

• Code https://github.com/InternLM/WildClawBench
• Dataset https://huggingface.co/datasets/internlm/WildClawBench

Which sources cite WildClawBench?

• WildClawBench

Related 智能体, Agent, 代码, Code, 创作, Creation, Agent Evaluation, In-the-Wild Tasks, Multimodal & Tool-Use Reasoning, 代码工程, 语言生成, Generation, 任务执行, Task Execution, 跨模态推理, Cross-modal Reasoning, 不支持, Unsupported benchmarks

• A-Bench
• A-OKVQA
• AbsPyramid
• ActionAtlas
• Ada-LEval
• AECBench
• AFQMC
• AgentBoard