Expired50% confidenceFactExact time
Alibaba Tongyi Lab released the agent evaluation benchmark Powbench V1.0, which contains 150 real-world tasks and 4,050 test units, with the combination of Quant 3.6 Max Preview and QuantPow achieving first place overall.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Related Claims
UnverifiedEleutherAI的Pythia项目同时公开了训练数据、代码和完整检查点序列，包含从70M到12B参数的8个规模梯度模型，全部在相同的Pile数据集上以相同顺序训练，并公开了超过150个检查点67% similarUnverifiedCodex使用了126个子智能体、32500次工具调用、391次浏览器测试和2300多次单元测试；Claude Code使用35个子智能体、102次浏览器测试和296次单元测试66% similarUnverifiedThe tested task was conducted over four rounds, dispatching a cumulative total of 68 sub-agents to complete batch prompt testing for a digital human writing skills module.65% similarUnverifiedPonytail's official benchmark covers 12 functional tasks and is run using Claude Code.65% similarUnverified在测试中向智能体输入「计算 6 乘 7」的请求，模型自动调用 ai_python_executable 工具并返回结果 42，整个推理过程在 MLflow 中完整记录65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/53944API
curl https://kongchang.com/api/v1/knowledge/claims/53944MCP
get_claim(id=53944)