// Xiaomi MiMo · Text LLM
MiMo-V2-Flash: VRAM & local deployment
309B total · 15B active (MoE) parameters. About 207 GB at 4-bit and 680 GB at BF16.
released 2025-12-16licence: Commercial OKmemory figures are estimates
01Specs
| Publisher | Xiaomi · XiaomiMiMo |
|---|---|
| Family / series | Xiaomi MiMo · MiMo-Flash |
| Type | Text LLM (text-generation) |
| Parameters | 309B total · 15B active (MoE) Source: official model card“309B total parameters and 15B active parameters” |
| Context length | 256K tokens (262,144)max_position_embeddings in config.json |
| Architecture | 48 layers · GQA · MoE |
| Released weights | FP8 (E4M3) · 313 GBdownload |
| Licence | MIT · Commercial OKPermissive licence (Apache-2.0 / MIT / CC-BY / OpenMDW): commercial use allowed. |
| Released | 2025-12-16 (Hugging Face repo created) |
02How much VRAM it needs (estimate)
Total = weights + 10% + 0.5GB overhead + KV cache for one request (FP16 KV). More concurrent users multiply the KV column.
| Precision | Weights | Smallest GPU | Mac |
|---|---|---|---|
| BF16 / FP16 (original) | 618 GB | 8 × H200 (1,128GB) | — |
| FP8 / INT8 / Q8_0 | 328 GB | 8 × H100 (640GB) | — |
| 4-bit (Q4_K_M / AWQ / GPTQ-Int4) | 187 GB | 4 × RTX PRO 6000 (384GB) | Mac 512GB |
KV cache not estimated: this model uses a new compressed / sparse attention design that our formula does not cover. Totals above are weights + overhead only.
Bits per weight: BF16 16, FP8 8.5, 4-bit 4.85. “Fits” = total ≤ 90% of memory; Mac GPU can use ~2/3 of unified memory (≤36GB) or ~3/4 (larger). For MoE, all experts must be in memory, so size by total parameters. Method →
03How to run it
Only runtimes the publisher or the runtime’s official library names for this model.
- SGLang — mentioned in the official model card
- Transformers — mentioned in the official model card
- KTransformers — mentioned in the official model card
04Official repositories & quantised builds
2 official repo(s) from XiaomiMiMo.
- MiMo-V2-Flash main
- MiMo-V2-Flash-Base Base
05Other Xiaomi MiMo models
All 14 →| Model | Params | 4-bit total | Released |
|---|---|---|---|
| MiMo-V2.6-Flash-MOPD | 309B / 15B | 207 GB | 2026-09-27 |
| MiMo-V2.6-Pro-MOPD | 1.02T / 42B | 681 GB | 2026-09-27 |
| MiMo-V2.6-Distill-Qwen-9B | 9.41B | 7.0 GB | 2026-09-21 |
| MiMo-V2.6-Flash-RL | 309B / 15B | 207 GB | 2026-09-21 |
| MiMo-V2.6-Pro-RL | 1.02T / 42B | 681 GB | 2026-09-21 |
| MiMo-V2.5 | 310B / 15B | 207 GB | 2026-04-27 |
| MiMo-V2.5-Pro | 1.02T / 42B | 681 GB | 2026-04-27 |
| MiMo-V2.5-ASR | 7.62B | 5.6 GB | 2026-04-23 |
06Sources
Data checked 2026-07-09. Parameters, context and licence are copied from the official repository; memory figures are our estimates.