OpenBMB· @OpenBMB · X·· 2026-09-09AI 评分37
AI 导读
面壁智能 MiniCPM5-2B 在 Q4_K_M 量化下以 23.0 秒完成 3/3 项真实任务，比 Qwen3.5-4B 快 34%，且成功率相同。测试涵盖播放列表构建、日志故障追踪与工具调用补货，Gemma 4 E2B 仅完成 2/3 任务。更快的本地推理让端侧智能体负载更具可行性。
正文 · 原文
Same Q4_K_M quant, same tasks, same success rate, MiniCPM5-2B can get tasks done quicker. This makes local agent workloads a lot more interesting. Thanks for the comparison @atomic_chat_hq
MiniCPM5-2B beats Qwen3.5-4B in 34% less time 👾 We gave three models the same real-world tasks and compared their attempts using Q4_K_M quants Tests: • Build a playlist within strict timing rules • Trace a checkout failure through server logs • Check stock and draft a replacement with tool calls Outputs: • MiniCPM5-2B: 3/3 tasks, 23.0s total • Qwen3.5-4B: 3/3 tasks, 35.0s total • Gemma 4 E2B: 2/3 tasks, 29.5s total MiniCPM and Qwen cleared all three, but MiniCPM finished faster, meanwhile Gemma found the failure’s cause but broke the JSON-only rule在 X 查看被引用的帖子
来源：OpenBMB · x.com