Skip to content

MiMo V2.6 Pro VRAM requirements

How much VRAM MiMo V2.6 Pro needs
FP4 + FP8 weights: 527 GBWith 8K context: 529 GBWith 32K context: 530 GB
Parameters
1T
Active per token
42B
Longest context
1M
Released
Sep 2026
License
MIT
Strongest at
Chat, Coding, Agents, Docs

How much VRAM MiMo V2.6 Pro needs

Memory MiMo V2.6 Pro needs at each precision and context length
PrecisionWeightsWith 8KWith 32KWith 128K
As released (FP4 + FP8)527 GB529 GB530 GB534 GB
3-bit466 GB468 GB469 GB473 GB
2-bit358 GB359 GB360 GB365 GB
1.6-bit197 GB198 GB199 GB204 GB

One conversation, 16-bit context cache, runtime overhead included.

GPUs and Macs that run MiMo V2.6 Pro

The best way MiMo V2.6 Pro runs on common GPUs, Macs and AI boxes
MachineBest fitContextSpeedFirst replyReport
RTX 4060Too bigPDF: MiMo V2.6 Pro on RTX 4060 report
RTX 5060Too bigPDF: MiMo V2.6 Pro on RTX 5060 report
RTX 3070Too bigPDF: MiMo V2.6 Pro on RTX 3070 report
RTX 3080 10 GBToo bigPDF: MiMo V2.6 Pro on RTX 3080 10 GB report
RTX 3060 12 GBToo bigPDF: MiMo V2.6 Pro on RTX 3060 12 GB report
Arc B580Too bigPDF: MiMo V2.6 Pro on Arc B580 report
RTX 4070Too bigPDF: MiMo V2.6 Pro on RTX 4070 report
RTX 4070 SuperToo bigPDF: MiMo V2.6 Pro on RTX 4070 Super report
RTX 5070Too bigPDF: MiMo V2.6 Pro on RTX 5070 report
RTX 4060 Ti 16 GBToo bigPDF: MiMo V2.6 Pro on RTX 4060 Ti 16 GB report
RTX 5060 Ti 16 GBToo bigPDF: MiMo V2.6 Pro on RTX 5060 Ti 16 GB report
Arc A770 16 GBToo bigPDF: MiMo V2.6 Pro on Arc A770 16 GB report
Radeon RX 9070 XTToo bigPDF: MiMo V2.6 Pro on Radeon RX 9070 XT report
RTX 4070 Ti SuperToo bigPDF: MiMo V2.6 Pro on RTX 4070 Ti Super report
RTX 4080 SuperToo bigPDF: MiMo V2.6 Pro on RTX 4080 Super report
RTX 5070 TiToo bigPDF: MiMo V2.6 Pro on RTX 5070 Ti report
RTX 5080Too bigPDF: MiMo V2.6 Pro on RTX 5080 report
Radeon RX 7900 XTXToo bigPDF: MiMo V2.6 Pro on Radeon RX 7900 XTX report
RTX 3090Too bigPDF: MiMo V2.6 Pro on RTX 3090 report
RTX 4090Too bigPDF: MiMo V2.6 Pro on RTX 4090 report
Radeon AI PRO R9700Too bigPDF: MiMo V2.6 Pro on Radeon AI PRO R9700 report
RTX 5090Too bigPDF: MiMo V2.6 Pro on RTX 5090 report
2× RTX 3090Too bigPDF: MiMo V2.6 Pro on 2× RTX 3090 report
2× RTX 4090Too bigPDF: MiMo V2.6 Pro on 2× RTX 4090 report
RTX PRO 5000 Blackwell 48 GBToo bigPDF: MiMo V2.6 Pro on RTX PRO 5000 Blackwell 48 GB report
RTX 6000 AdaToo bigPDF: MiMo V2.6 Pro on RTX 6000 Ada report
L40SToo bigPDF: MiMo V2.6 Pro on L40S report
2× RTX 5090Too bigPDF: MiMo V2.6 Pro on 2× RTX 5090 report
4× RTX 3090Too bigPDF: MiMo V2.6 Pro on 4× RTX 3090 report
RTX PRO 6000 BlackwellToo bigPDF: MiMo V2.6 Pro on RTX PRO 6000 Blackwell report
Mac mini M4 16 GBToo bigPDF: MiMo V2.6 Pro on Mac mini M4 16 GB report
Mac mini M4 24 GBToo bigPDF: MiMo V2.6 Pro on Mac mini M4 24 GB report
Mac mini M4 32 GBToo bigPDF: MiMo V2.6 Pro on Mac mini M4 32 GB report
Mac mini M4 Pro 24 GBToo bigPDF: MiMo V2.6 Pro on Mac mini M4 Pro 24 GB report
Mac mini M4 Pro 48 GBToo bigPDF: MiMo V2.6 Pro on Mac mini M4 Pro 48 GB report
Mac mini M4 Pro 64 GBToo bigPDF: MiMo V2.6 Pro on Mac mini M4 Pro 64 GB report
M4 Max 128 GBToo bigPDF: MiMo V2.6 Pro on M4 Max 128 GB report
M5 Max 128 GBToo bigPDF: MiMo V2.6 Pro on M5 Max 128 GB report
M3 Ultra 96 GBToo bigPDF: MiMo V2.6 Pro on M3 Ultra 96 GB report
M3 Ultra 256 GBToo bigPDF: MiMo V2.6 Pro on M3 Ultra 256 GB report
M3 Ultra 512 GB2-bit128K25 tok/s26 sPDF: MiMo V2.6 Pro on M3 Ultra 512 GB report
M5 Ultra 512 GB2-bit128K33 tok/s8.1 sPDF: MiMo V2.6 Pro on M5 Ultra 512 GB report
Ryzen AI Max+ 395 64 GBToo bigPDF: MiMo V2.6 Pro on Ryzen AI Max+ 395 64 GB report
Ryzen AI Max+ 395 128 GBToo bigPDF: MiMo V2.6 Pro on Ryzen AI Max+ 395 128 GB report
RTX SparkToo bigPDF: MiMo V2.6 Pro on RTX Spark report
DGX SparkToo bigPDF: MiMo V2.6 Pro on DGX Spark report
Dual DGX Spark1.6-bit608K20 tok/s5.3 sPDF: MiMo V2.6 Pro on Dual DGX Spark report
DGX Station GB3001.6-bit832K168 tok/s0.7 sPDF: MiMo V2.6 Pro on DGX Station GB300 report
2× RTX PRO 6000 BlackwellToo bigPDF: MiMo V2.6 Pro on 2× RTX PRO 6000 Blackwell report
4× RTX PRO 6000 Blackwell2-bit80K60 tok/s0.7 sPDF: MiMo V2.6 Pro on 4× RTX PRO 6000 Blackwell report
8× RTX PRO 6000 BlackwellFP4 + FP81M33 tok/s0.4 sPDF: MiMo V2.6 Pro on 8× RTX PRO 6000 Blackwell report
A100 80 GB SXMToo bigPDF: MiMo V2.6 Pro on A100 80 GB SXM report
H100 SXMToo bigPDF: MiMo V2.6 Pro on H100 SXM report
4× H200 NVLFP4 + FP8112K71 tok/s0.7 sPDF: MiMo V2.6 Pro on 4× H200 NVL report
8× H200 NVLFP4 + FP81M71 tok/s0.4 sPDF: MiMo V2.6 Pro on 8× H200 NVL report

Best fit is the highest quality, up to 8-bit, that leaves 16K of context. First reply is the wait for a 2K prompt. Speeds are rough, plus or minus 25%.

How to run it

llama.cpp
llama-server -hf ggml-org/MiMo-V2.6-Pro-RL-GGUF
vLLM
vllm serve XiaomiMiMo/MiMo-V2.6-Pro-RL
MLX (Mac)
mlx_lm.chat --model mlx-community/MiMo-V2.6-Pro-RL-mxfp4-q8

Questions

How much VRAM does MiMo V2.6 Pro need?

MiMo V2.6 Pro needs 527 GB for its weights as released (FP4 + FP8), 529 GB with an 8K context and 530 GB with 32K.

What is the minimum VRAM for MiMo V2.6 Pro?

About 198 GB at 1.6-bit with an 8K context, the lowest precision we recommend for a model this size. The smallest machine here that runs it without spilling is the M3 Ultra 512 GB.

How big is MiMo V2.6 Pro?

1T parameters (42B active per token): 527 GB of weights as released (FP4 + FP8).

Can MiMo V2.6 Pro run on an RTX 4090?

Not on one card. The smallest machine here that runs it is the M3 Ultra 512 GB.

More models