FITMYLLM · NOVEMBER 4, 2026 · OPEN METHODOLOGY · COMMUNITY BENCHMARKS
LIVE · UPDATED DAILY
310.76B
Parameters (15B active)
Quantization Options Select your GPU for speed estimates Context length: 4K 8K 16K 32K 64K 128K 256K 512K 1M
Quant Bits VRAM @ 16K Quality IQ2_XXS 2.38 93.3 GB
92.9 + 0.3 KV
low IQ2_M 2.93 114.6 GB
114.3 + 0.3 KV
low Q2_K 3.16 123.6 GB
123.2 + 0.3 KV
low IQ3_XXS 3.25 127.1 GB
126.7 + 0.3 KV
low IQ3_XS 3.5 136.8 GB
136.4 + 0.3 KV
low Q3_K_S 3.64 142.2 GB
141.9 + 0.3 KV
low IQ3_M 3.76 146.9 GB
146.5 + 0.3 KV
low Q3_K_M 4 156.2 GB
155.9 + 0.3 KV
low Q3_K_L 4.3 167.9 GB
167.5 + 0.3 KV
moderate IQ4_XS 4.46 174.1 GB
173.7 + 0.3 KV
moderate Q4_K_S 4.67 182.2 GB
181.9 + 0.3 KV
moderate Q4_K_M 4.89 190.8 GB
190.4 + 0.3 KV
good Q5_K_S 5.57 217.2 GB
216.9 + 0.3 KV
good Q5_K_M 5.7 222.2 GB
221.9 + 0.3 KV
good Q6_K 6.56 255.6 GB
255.3 + 0.3 KV
excellent Q8_0 8.5 331.0 GB
330.7 + 0.3 KV
lossless FP16 16 622.3 GB
622.0 + 0.3 KV
lossless
Select your GPU above to see speed estimates and compatibility for each quantization.
Too big for a single GPU — plan a multi-GPU deployment
Even the lightest quant needs ~93 GB. Size GPUs, replicas, TCO and scaling for a production setup. Open in Enterprise →
▸ READY TO RUN THIS? RENT BY THE HOUR
RENT A GPU AND RUN MIMO V2.6 FLASH NOW
Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.
Community Ratings Chat Coding Reasoning Creative Vision Roleplay Agentic
Loading ratings...
Run this model IQ2_XXS — 92.9 GB VRAM IQ2_M — 114.3 GB VRAM Q2_K — 123.2 GB VRAM IQ3_XXS — 126.7 GB VRAM IQ3_XS — 136.4 GB VRAM Q3_K_S — 141.9 GB VRAM IQ3_M — 146.5 GB VRAM Q3_K_M — 155.9 GB VRAM Q3_K_L — 167.5 GB VRAM IQ4_XS — 173.7 GB VRAM Q4_K_S — 181.9 GB VRAM Q4_K_M — 190.4 GB VRAM Q5_K_S — 216.9 GB VRAM Q5_K_M — 221.9 GB VRAM Q6_K — 255.3 GB VRAM Q8_0 — 330.7 GB VRAM FP16 — 622.0 GB VRAM
Ollama llama.cpp vLLM LM Studio KoboldCpp Jan Docker
▸ Easiest way to get started · Beginners
DOCS ↗ curl -fsSL https://ollama.com/install.sh | shCOPY
$ ollama run mimo:311b-q4_K_MCOPY
Tag may need adjustment — check ollama.com/library/mimo for available tags.
▸ SETUP GUIDE >_
Auto-setup with fitmyllm CLI Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.
Auto-detect GPU Live tok/s in chat Speed benchmarks 9 inference engines
GPUs that can run this model At Q4_K_M quantization. Sorted by minimum VRAM.
Find the best GPU for MiMo V2.6 Flash
Build Hardware for MiMo V2.6 Flash ▸ COLOPHON FITMYLLM · INDEPENDENT · DATA-DRIVEN
FITMYLLM · EST. 2025 · © 2026
RECOMMENDATIONS FROM PUBLISHED MATH, CORRECTED BY THE COMMUNITY — 30.
▸ SPEC SHEET
MiMo V2.6 Flash — 310.76B MoE. ▸ SPECIFICATIONS
PARAMETERS 310.76B (15B active)
ARCHITECTURE Mixture of Experts
CONTEXT LENGTH 1024K tokens
CAPABILITIES chat, coding, reasoning, multilingual, vision, audio, math, agentic, tool_use
RELEASE DATE 2026-09-21
PROVIDER Xiaomi
FAMILY mimo ▸ VRAM REQUIREMENTS
QUANT BPW VRAM QUALITY IQ2_XXS 2.38 92.9 GB 65% IQ2_M 2.93 114.3 GB 75% Q2_K 3.16 123.2 GB 78% IQ3_XXS 3.25 126.7 GB 82% IQ3_XS 3.5 136.4 GB 84% Q3_K_S 3.64 141.9 GB 85% IQ3_M 3.76 146.5 GB 86% Q3_K_M 4 155.9 GB 88% Q3_K_L 4.3 167.5 GB 90% IQ4_XS 4.46 173.7 GB 92% Q4_K_S 4.67 181.9 GB 93% Q4_K_M 4.89 190.4 GB 94% Q5_K_S 5.57 216.9 GB 96% Q5_K_M 5.7 221.9 GB 96% Q6_K 6.56 255.3 GB 97% Q8_0 8.5 330.7 GB 100% FP16 16 622.0 GB 100%
§ 01 BENCHMARK SCORES
HLE 35.1
aa_scicode 51.3
aa_lcr 74.3
aa_terminal_bench_v4 22.7
§ 03 COMPATIBLE GPUs
13 @ Q4_K_M Feedback