MiMo-V2.5-Pro 1T — 显存、速度与本地部署
作者： Jakub Rusinowski · 最后更新：
The flagship: 1 trillion total, 42B active per token, MIT, 1M context. Matches frontier coding models on published comparisons at a licence that imposes no revenue threshold. Datacenter-class to self-host at about 605 GB at Q4_K_M.
MiMo-V2.5-Pro 1T 在 Q4_K_M 下约需 605 GB 显存——量化权重加框架开销，不含 KV 缓存。在 Apple Silicon 上，这部分来自统一内存。
按量化级别的显存与速度
计算基准：NVIDIA RTX 4090 (24 GB)。仅含权重与开销：该模型架构未公开，因此未计入 KV 缓存。
| 量化 | 显存 | 显存 | 速度（估算） | 适配 |
|---|---|---|---|---|
| Q2_K 2.63 bpw | 329.6 GB | — | 放不下 | |
| Q3_K_M 3.41 bpw | 427.1 GB | — | 放不下 | |
| Q4_K_M 4.83 bpw | 604.6 GB | — | 放不下 | |
| Q5_K_M 5.67 bpw | 709.6 GB | — | 放不下 | |
| Q6_K 6.56 bpw | 820.8 GB | — | 放不下 | |
| Q8_0 8.50 bpw | 1063.3 GB | — | 放不下 | |
| F16 16.00 bpw | 2000.8 GB | — | 放不下 |
黑色标记 = NVIDIA RTX 4090 (24 GB) 上的可用显存。 估算来自内存带宽屋顶线模型，详见 方法说明页. MiMo-V2.5-Pro 1T 显存计算器 →
运行 MiMo-V2.5-Pro 1T
MiMo-V2.5-Pro 1T is not practical on a single consumer GPU: it needs about 604.6 GB at Q4_K_M, more than any single device in our catalogue holds with headroom. How multi-GPU inference works, or rent a multi-GPU server by the hour:
如何运行 MiMo-V2.5-Pro 1T
安装 Ollama，然后运行：
ollama run mimo-v2-5规格
Verified — Checked against the primary source — the model card or the vendor spec page — and corroborated by a second independent source.
- 参数量
- 1 Trillion (42B active)
- 上下文窗口
- 1,000,000
- 架构
- Mixture-of-Experts
- 提供商
- Xiaomi
- 许可证
- MIT
- Practical deployment
- Private cloud / datacenter cluster
- 规格量化
- Q4_K_M
- 系统内存
- 768 GB
- 记录更新于
- 2026-09-06
Commercial use permitted. No usage restrictions beyond attribution.
另请参阅 MiMo-V2.6-Pro
质量与使用场景
评分由模型作者或独立评测方发布——衡量质量而非吞吐量，并非我们实测。
我的 GPU 能运行 MiMo-V2.5-Pro 1T 吗？
- MiMo-V2.5-Pro 1T 在 AMD Ryzen AI Max+ 395 上
- MiMo-V2.5-Pro 1T 在 Apple M1 Ultra 上
- MiMo-V2.5-Pro 1T 在 Apple M2 Max 上
- MiMo-V2.5-Pro 1T 在 Apple M2 Ultra 上
- MiMo-V2.5-Pro 1T 在 Apple M4 Max 上
- MiMo-V2.5-Pro 1T 在 Apple M5 Max 上
- MiMo-V2.5-Pro 1T 在 NVIDIA A100 80GB (PCIe) 上
- MiMo-V2.5-Pro 1T 在 NVIDIA DGX Spark 上
- MiMo-V2.5-Pro 1T 在 NVIDIA H100 80GB (PCIe) 上
- MiMo-V2.5-Pro 1T 在 NVIDIA RTX PRO 6000 Blackwell 上
MiMo-V2.5 的其他尺寸
MiMo-V2.5-Pro 1T — 常见问题
How much VRAM does MiMo-V2.5-Pro 1T need?
About 605 GB at Q4_K_M — quantized weights plus framework overhead, before any KV cache. The cache grows with context length and is added on top; the table above folds it in. Apple Silicon counts unified memory toward the same figure.
Does MiMo-V2.5-Pro 1T run on an RTX 4090 (24 GB)?
No. MiMo-V2.5-Pro 1T needs about 605 GB at Q4_K_M, more than a single RTX 4090's 24 GB. It needs a larger card, several GPUs, or Apple Silicon with enough unified memory — or it runs with part of the weights offloaded to system RAM, which is much slower.
How do I run MiMo-V2.5-Pro 1T locally?
Install Ollama and run `ollama run mimo-v2-5`. That pulls the weights and starts a local OpenAI-compatible endpoint; after the download nothing leaves the machine.
What other sizes does MiMo-V2.5 come in?
MiMo-V2.5 310B (188 GB), MiMo-V2.5-Pro 1T (605 GB). Every size shares the family's training and licence; the larger ones score higher and need proportionally more memory.