PulseAugur
中
实时 15:10:17
English(EN) Mistral Large 4 on your own GPUs: does 1 trillion parameters fit on 8 H200s, and what does the license allow?

Mistral Large 4 开源权重发布,需要大量 GPU 资源

Mistral AI 发布了 Mistral Large 4 的开源权重,这是一个多模态模型,拥有约 1 万亿参数和 100 万个 token 的上下文窗口。仅 FP8 检查点就需要超过 1 TB 的存储空间,这使得它勉强能装入 8 个 H200 GPU,仅权重就会占用 94% 的容量。随附的许可证允许免费的内部使用,但禁止转售,并要求为“模型即服务”产品单独获得商业许可证。 AI

影响 需要大量的 GPU 基础设施,可能限制许多组织自行托管。

排序理由 Frontier-lab 模型发布,包含开源权重和系统卡。[lever_c_demoted from frontier_release: ic=1 ai=1.0]

在 dev.to — LLM tag 阅读 →

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

Mistral Large 4 开源权重发布,需要大量 GPU 资源

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Significant
Frontier-lab 模型发布,包含开源权重和系统卡。[lever_c_demoted from frontier_release: 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, infra
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
11 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. dev.to — LLM tag TIER_1 English(EN) · ai maya ·

    Mistral Large 4 可在您自己的 GPU 上运行:1 万亿参数是否能装入 8 个 H200,以及许可协议允许什么?

    <h2> TL;DR </h2> <p>Mistral published the open weights of Mistral Large 4 on 2026-10-29: about 1 trillion parameters in total, 49 billion active per token, a 1M-token context window, and a modified MIT license. I did not run the model. I read the public <code>config.json</code>, …