English(EN) ModelBest Open-Sources MiniCPM-V 4.7, Its First MoE Multimodal Model, Running in 3GB on a Phone

ModelBest 开源 MiniCPM-V 4.7 多模态 MoE 模型

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-14 04:24

ModelBest 已开源其首个混合专家（MoE）多模态模型 MiniCPM-V 4.7。该模型提供 35B-A3B 和 1B 两种尺寸。据报道，较大的 MoE 版本可以在内存约 3GB 的智能手机上运行，解码速度超过每秒 13 个 token。 AI

影响以低内存需求实现设备端多模态 AI 功能。

排序理由非前沿实验室发布新模型开源。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 23 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 非前沿实验室发布新模型开源。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: model release, product; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

Pandaily TIER_1 English(EN) · [email protected] (Pandaily) · 2026-10-14 04:24

ModelBest 开源 MiniCPM-V 4.7，首个手机上 3GB 运行的 MoE 多模态模型

ModelBest's MiniCPM-V 4.7 comes in 35B-A3B and 1B sizes. The MoE version decodes at over 13 tokens/s in about 3GB of memory on an iPhone 17 Pro Max, the company says.