LocalForge
返回指南
模型 / COMMUNITY REAP BUILD

Qwen3.8 Flash Next REAP-288

模型卡核对于 2026-09-16 · 内存适配为估算值；除非有与你的配置相符的独立实测，否则速度未知。

也会被搜索为：qwen3.8-flash-next-reap-288-mlx-4bit

Community-pruned conversion, not the stock checkpoint. The 39 GB model / ~40.57 GB total resident path requires oMLX 0.6.4+ and streams only the sparse PLE table from NVMe. A published M4 Pro 48 GB test measured 26–29 tok/s and recommended 7,680 tokens; do not generalize it to arbitrary models or Macs.

一手来源
总参数176B
估算内存~40.57 GBREAP-288 · MLX 4-bit · PLE on NVMe
上下文7.6K recommended on measured 48 GB MacText
许可证Qwen CommunityMultilingual; post-pruning quality not independently established
01 / 适合
  • 48 GB Mac experiments
  • coding agents
  • model-specific NVMe streaming
02 / 估算适配
Mac Studio M5 Ultra · 96 GB余量充足MacBook Pro M5 Max · 64 GB勉强适配Mac Studio M3 Ultra · 96 GB余量充足Mac Studio M3 Ultra · 256 GB余量充足Apple silicon · 64 GB勉强适配
03 / 部署方式

下面是运行方式示例，不是已验证的安装流程。容量估算对应量化文件，官方仓库可能是全精度权重；MLX、GGUF 与 TensorRT 需要匹配的转换文件或引擎。

oMLX进阶omlx serve --model-dir ~/models --memory-guard-gb 48
注意

Community-pruned conversion, not the stock checkpoint. The 39 GB model / ~40.57 GB total resident path requires oMLX 0.6.4+ and streams only the sparse PLE table from NVMe. A published M4 Pro 48 GB test measured 26–29 tok/s and recommended 7,680 tokens; do not generalize it to arbitrary models or Macs.