Ministral 3 — Mistral AI 的本地 AI 模型

作者: Jakub Rusinowski · 最后更新:

The edge tier of the Mistral 3 family — 3B, 8B and 14B dense, every size shipped in base, instruct and reasoning variants with image understanding, all Apache 2.0. Built for laptops, phones, drones and anywhere a datacenter is not. The 3B runs in about 4 GB of video memory; the 14B reasoning variant scores 85% on AIME 2025.

变体

Ministral 3 最小的变体在 Q4_K_M 下约需 3 GB 显存——量化权重加框架开销,不含 KV 缓存。

模型显存
Ministral 3 3B →
3B
~2.6 GB
Ministral 3 8B →
8B
~5.6 GB
Ministral 3 14B →
14B
~9.3 GB

显存为 Q4_K_M 下的量化权重加开销,与 GPU 与显存检测器使用同一引擎计算。

如何在本地运行 Ministral 3

安装 Ollama,然后拉取标签。

ollama run ministral-3:3b

在上方选择一个尺寸,查看它自己的显存、速度估算和安装命令。

许可证

Apache-2.0允许商业使用

Commercial use permitted. No usage restrictions beyond attribution.

适用于: Ministral 3 3B, Ministral 3 8B, Ministral 3 14B

推荐 GPU

目录中能在本地运行 Ministral 3(至少 3 GB 显存)的最便宜 GPU 是 Intel Arc B570 (10 GB).

联盟营销声明: 本页部分链接为联盟推广链接——如果你通过它们购买,LLM Configurator 可能会获得佣金,而你无需支付任何额外费用。作为亚马逊联盟成员(Amazon Associate),LLM Configurator 会从符合条件的购买中获得收益。
Intel Arc B570 10GB
10 GB VRAM · 150 W board power
2026年价格波动较大——请以当前商品页价格为准。

我的 GPU 能运行 Ministral 3 吗?

Ministral 3 — 常见问题

How much VRAM does Ministral 3 need?

Ministral 3 needs about 3 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Ministral 3 3B (3 GB, Q4_K_M); Ministral 3 8B (6 GB, Q4_K_M); Ministral 3 14B (9 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.

Can I run Ministral 3 on an RTX 4090 (24 GB)?

Yes — Ministral 3 runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.

What quantization should I use for Ministral 3?

Q4_K_M is the best balance of quality and VRAM for Ministral 3 in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.

How do I run Ministral 3 with Ollama?

Install Ollama, then run: ollama run ministral-3:3b. This downloads Ministral 3 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.