MiMo V2 Flash

Released 2025-12 · reasoning · 256K tokens · 2 benchmarks · Open weight

Editorial notes

Lanzado 16 diciembre 2025. MoE 309B totales / 15B activos. MIT license, open weights en HuggingFace. 256K context. $0.10/$0.30 per MTok. Predecesor del V2 Pro. AA Coding Index 33.5.

Spec sheet

Empresa: Xiaomi
Pais: CN
Tipo: reasoning
Release: 2025-12
Context: 256K tokens
Params total: 309B
Params active (MoE): 15B
Licencia: mit
Quants: BF16, Q8_0, Q5_K_M, Q4_K_M, Q3_K_M
Pricing (openrouter): $0.1/$0.3/M
Slug: mimo-v2-flash

Quick install

1 tools

ollama.com

ollama run mimo:v2-flash

Nota: Q4_K_M ~180 GB VRAM

Benchmarks (2)

Reasoning 2

Cite this model

BibTeX · APA

BibTeX

@misc{frontier-mimo-v2-flash,
  title  = {MiMo V2 Flash},
  author = {{Xiaomi}},
  year   = {2025},
  note   = {Frontier Benchmarks AI atlas. Accessed 2026-05-08},
  url    = {https://frontierbenchmarks.com/models/mimo-v2-flash}
}

APA

Xiaomi (2025). MiMo V2 Flash [Large language model]. Frontier Benchmarks AI. Retrieved 2026-05-08, from https://frontierbenchmarks.com/models/mimo-v2-flash

Citation refleja la pagina del atlas, no el paper original del modelo. Para el paper, ve a la seccion "Recursos" arriba.

⚔️ Battle vs otro modelo ← Todos los modelos Mas de Xiaomi Methodology