All models › xiaomi

MiMo-V2.6-Pro

xiaomi · MIT license · on Arena since 25 Sept 2026

Download
Arena score1491#11 of 416 in Chat, paid models included
Size1024.2 B44.9 B active per token (mixture of experts)
4-bit download621 GBweights only, Q4_K_M
Max context1024Ktokens, from the model config

Where to use it for free

Arena scores

ArenaScoreRankVotes
Chat1491#11 of 4164K
Coding1625#20 of 1445K
Vision1264#50 of 1652K

Rank among all models, paid ones included. Source: Arena leaderboard dataset.

Speed on popular hardware

HardwareGPU memoryFits?Speed
RTX 3060 12 GB12 GBDoes not fit—
RTX 4060 Ti 16 GB16 GBDoes not fit—
RTX 309024 GBDoes not fit—
RTX 409024 GBDoes not fit—
RTX 509032 GBDoes not fit—
RX 7900 XTX24 GBDoes not fit—
2 × RTX 309048 GBDoes not fit—
4 × RTX 309096 GBDoes not fit—
Mac M4 Max 128 GB96 GBDoes not fit—
Ryzen AI Max+ 395 128 GB96 GBDoes not fit—
8 × H100 SXM640 GBFits in GPU66 tok/s

Estimate at 4-bit with 32K context, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 70% ÷ data read per token; real numbers vary by about ±30%.

GPU memory needed

Level8K context32K context128K contextQuality vs full
Q3_K_M513 GB515 GB521 GB+4.7% noticeable loss
Q4_K_M622 GB623 GB629 GB+1.2% very small loss
Q5_K_M729 GB731 GB737 GB+0.4% near-lossless
Q6_K841 GB843 GB849 GB+0.2% near-lossless
Q8_01089 GB1091 GB1097 GB≈ 0% same as full
F162050 GB2051 GB2057 GB≈ 0% same as full

Weights + context cache (from the model's architecture) + runtime. Quality: increase in perplexity against the full model, from llama.cpp tests, scaled by model size. Simulate your own setup.

Chat score over time

MiMo-V2.6-Pro 1491Best paid model

Arena score from 17 Apr 2026 to 15 Oct 2026.

Vision score over time

MiMo-V2.6-Pro 1264Best paid model

Arena score from 10 Apr 2026 to 9 Oct 2026.

Similar free models in Chat