AI NEWS.ccAI正在发生 未来正在更新
♡ 我的收藏 0
AI 模型库/ddalcu/Qwen3.8-Flash-Next-MLX-Serve-mixed-4-8bit
MODEL PROFILE · 模型档案

Qwen3.8-Flash-Next-MLX-Serve-mixed-4-8bit

ddalcu · Qwen3.8-Flash-Next-MLX-Serve-mixed-4-8bit

> **Use Qwen3.8-Flash-Next-MLX-Serve-iQ-MLX-4.7bpw instead.** > It is this pack re-quantized with an importance matrix collected on the bf16 > model: same layout, same 75 GB, same bytes read per token, closer to bf16 > (KLD median 0.0052 vs 0.0066, top-1 88.6% vs 87.9% against bf16's own > logits). This pack stays up for existing installs but gets no more updates. > > Switching from this pack? `ngram_table.bin` (32 GB) is identical in both. > Move it into the new pack's folder before downloading and the download > skips it: > > mlx-serve pack of Qwen/Qwen3.8-Flash-Next, the Qwen4 preview architecture (`model_type: qwen4_exp`). Runs on a 128 GB Mac with about 75 GB resident. Includes the MTP head and the vision tower (image and video input). Measured on an M4 Max 128 GB with mlx-serve 26.8.11: ~75 GB resident, decode ~60 tok/s serial and 78 tok/s with MTP (`--mtp`; +41% on code, a few percent slower on prose), prefill ~730 tok/s, a needle at 24.8k tokens recovered with sparse attention engaged. Prefix cache on, images included.

译文排队中，当前展示来源原文。

文本生成license:other
上下文窗口来源未披露
输入 → 输出文本 → 文本
模型权重待核验
档案更新时间2026/10/30 08:42:25
OVERVIEW

模型介绍

> **Use Qwen3.8-Flash-Next-MLX-Serve-iQ-MLX-4.7bpw instead.** > It is this pack re-quantized with an importance matrix collected on the bf16 > model: same layout, same 75 GB, same bytes read per token, closer to bf16 > (KLD median 0.0052 vs 0.0066, top-1 88.6% vs 87.9% against bf16's own > logits). This pack stays up for existing installs but gets no more updates. > > Switching from this pack? `ngram_table.bin` (32 GB) is identical in both. > Move it into the new pack's folder before downloading and the download > skips it: > > mlx-serve pack of Qwen/Qwen3.8-Flash-Next, the Qwen4 preview architecture (`model_type: qwen4_exp`). Runs on a 128 GB Mac with about 75 GB resident. Includes the MTP head and the vision tower (image and video input). Measured on an M4 Max 128 GB with mlx-serve 26.8.11: ~75 GB resident, decode ~60 tok/s serial and 78 tok/s with MTP (`--mtp`; +41% on code, a few percent slower on prose), prefill ~730 tok/s, a needle at 24.8k tokens recovered with sparse attention engaged. Prefix cache on, images included.

主要特点

特点摘要待依据来源补齐，不凭模型名称推断能力。

局限与注意事项

暂无足够来源支持具体缺点结论。请重点验证实际任务表现、时延、数据政策、许可与版本生命周期。

PRICING & COST

来源报价与费用分析

同一模型在不同渠道、地区和配置下可能采用不同价格。下表独立保留各渠道报价，不将渠道价格视为厂商官方统一价格。

来源 / 渠道输入输出计费单位采集时间
暂未取得可核验报价，请查阅来源渠道。

按月估算基础 Token 费用

—

预估不包含缓存优惠、阶梯定价、多模态附加费用、网络检索、税费及推理额外输出。

INDEPENDENT BENCHMARKS

公开评测与运行配置

暂无已采集的可核验公开评分。模型未上榜不代表能力较低。

查看完整已收录榜单 →

常见问题

开放权重是否代表可以免费商用？

不一定。应核对具体版本许可证、商用门槛和使用政策。自部署还会产生算力和维护成本。

这些价格是实时结算价格吗？

不是实时结算价格，而是带采集时间的来源报价快照。请到对应渠道核对当前版本、区域、折扣和计费方式。

资料来源与可追溯记录

huggingface · ddalcu/Qwen3.8-Flash-Next-MLX-Serve-mixed-4-8bit ↗2026/10/29 21:12:34
huggingface · enrich:8eb4a749-26cb-400d-a68e-c10473a81332 ↗2026/10/30 08:42:25

许可证：other。模型介绍与第三方评测均以原来源的具体版本和方法为准。