MiMo-V2.5 — local AI model by Xiaomi
Written by Jakub Rusinowski · Last updated
Xiaomi's MIT-licensed MoE pair, and the clearest licence story at the trillion-parameter tier — MIT, not a bespoke open-weight agreement with revenue thresholds. Both carry 1M-token context and agentic multimodal capability. Consistently omitted from best-of roundups despite matching Kimi K2.6 on aggregate index scores.
Variants
The smallest MiMo-V2.5 variant needs about 188 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| MiMo-V2.5 310B → 310B/15B active | ~188 GB | 1,000,000 | ollama run mimo-v2-5 | |
| MiMo-V2.5-Pro 1T → 1T/42B active | ~604.6 GB | 1,000,000 | ollama run mimo-v2-5 |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run MiMo-V2.5 locally
Install Ollama, then pull the tag.
ollama run mimo-v2-5Pick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: MiMo-V2.5 310B, MiMo-V2.5-Pro 1TRecommended GPU
The cheapest catalogued GPU that runs MiMo-V2.5 locally (min 188 GB VRAM) is the Apple M3 Ultra (512 GB).
Can I run MiMo-V2.5 on my GPU?
- MiMo-V2.5 on AMD Ryzen AI Max+ 395
- MiMo-V2.5 on Apple M1 Ultra
- MiMo-V2.5 on Apple M2 Max
- MiMo-V2.5 on Apple M2 Ultra
- MiMo-V2.5 on Apple M4 Max
- MiMo-V2.5 on Apple M5 Max
- MiMo-V2.5 on NVIDIA A100 80GB (PCIe)
- MiMo-V2.5 on NVIDIA DGX Spark
- MiMo-V2.5 on NVIDIA H100 80GB (PCIe)
- MiMo-V2.5 on NVIDIA RTX PRO 6000 Blackwell
MiMo-V2.5 — frequently asked questions
How much VRAM does MiMo-V2.5 need?
MiMo-V2.5 needs about 188 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: MiMo-V2.5 310B (188 GB, Q4_K_M); MiMo-V2.5-Pro 1T (605 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run MiMo-V2.5 on an RTX 4090 (24 GB)?
MiMo-V2.5's smallest variant needs about 188 GB, which exceeds a single RTX 4090 (24 GB). Use multiple GPUs, a higher-VRAM card, or Apple Silicon with large unified memory.
What quantization should I use for MiMo-V2.5?
Q4_K_M is the best balance of quality and VRAM for MiMo-V2.5 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 MiMo-V2.5 with Ollama?
MiMo-V2.5 has no local Ollama tag — the published tag is cloud-hosted, so running it sends your prompts to a hosted GPU rather than your own machine.