Unverified50% confidenceBenchmarkExact time
Under the same Agent framework and prompt, Qwen3 27B (64K context) outperformed Leizhen T1 and Yizhentuo-Ri (both with 256K context) in PowerPoint creation.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/1/2026
First Seen
Valid until: 11/30/2026
Sources
Related Entities
Related Claims
UnverifiedAfter completing the PowerPoint, Qwen3 27B called a vision model to self-review, achieving a generate–review–revise loop, ultimately producing a 16-page fully element-editable PowerPoint presentation.68% similarUnverifiedOpen-source models such as Qwen-7B and LLaMA-2-7B, after Agent Tuning, can achieve tool-calling accuracy at the level of GPT-3.5 or even close to GPT-467% similarUnverifiedQwen2.5-27B是阿里云通义实验室发布的270亿参数开源模型，在MMLU、HumanEval、GSM8K等基准上接近GPT-4早期版本水平63% similarUnverifiedQwen2.5-Coder 在 HumanEval 等代码生成基准上展现出接近 GPT-4o 的表现，其7B版本量化后可在普通笔记本上运行63% similarUnverifiedShotcut在MLT之上使用Qt框架构建图形界面，结合现代C++编写业务逻辑61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/837546API
curl https://kongchang.com/api/v1/knowledge/claims/837546MCP
get_claim(id=837546)