XiaomiMiMoV2ForCausalLM

MiMo V2.5 Pro — Hardware Requirements & GPU Compatibility

ChatFunctionsCode

MiMo V2.5 Pro is a 1023.2B-parameter open language model from Xiaomi. It supports a context window of up to 1,048,576 tokens. At Q4_K_M it needs about 614.47 GB of VRAM — see which GPUs and Macs can run it below.

1023.2B params · MoE44B active1049K contextMITApr 202619.6K downloads753 likes4.1K quant downloads

Your hardware

Verdicts below are for the reference RTX 4090. Set yours to personalise the page.

Detection reads your browser's GPU name with no data sent anywhere.

MiMo V2.5 Pro runs on a multi-GPU server

Q4_K_M needs ~616 GB · about 5× H200 SXM 141 GB · or Mac Studio (M3 Ultra, 512GB) at Q3_K_M (~20 tok/s)

Macs from 512 GB (Q3_K_M · ~20 tok/s)

Score 116.4#17 of 100 open#14 of 96 that fit ≤640 GBMIT

Not set up for it?

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

llmrun Composite Score →

116.4

90% range 111.3–121.6
#17 of 100 open models
#14 of 96 that fit ≤640 GB

Benchmarks · tier chart ↓

Among 100 open modelsHow the Composite Score is built

Coding115.690% range 106–126
AgentsNot enough data yet
MathNot enough data yet
Science114.690% range 97–133
Reasoning110.890% range 102–120

Run MiMo V2.5 Pro

One recommended way. Every other method, OS and quant is in the step-by-step guide.

Other ways

each opens step-by-step

Run in cloud

Rent 5× H200 SXM · Q4_K_M

Estimated hourly price: ~$17.95/hr; est., whole server
GPUs: 5 GPUs; 705 GB of VRAM

Deployon RunPodRent H200 SXM$3.59/hr per GPU

Compare GPUs →

How Much VRAM Does MiMo V2.5 Pro Need?

Select a quantization level; the listed hardware adheres to it.

VRAM	+32K	+128K	File	Quality	Runs on
281.9 GB	285.2 GB	295.8 GB	317.2 GB	2-bit · i-quant extreme compression — Importance-weighted 2-bit, extreme compression — significant quality loss	Runs on Instinct MI350X 288 GB
320.3 GB	323.6 GB	334.2 GB	319.4 GB	2-bit · i-quant small — Importance-weighted 2-bit, small	Runs on 5× H100 SXM 80 GB
345.9 GB	349.2 GB	359.7 GB	317.3 GB	2-bit · i-quant medium — Importance-weighted 2-bit, medium	Runs on 5× H100 SXM 80 GB
397.0 GB	400.3 GB	410.9 GB	412.7 GB	3-bit · i-quant — Importance-weighted 3-bit	Runs on 6× H100 SXM 80 GB
435.4 GB	438.7 GB	449.3 GB	~434.9 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Runs on 7× H100 SXM 80 GB
435.4 GB	438.7 GB	449.3 GB	377.9 GB	3-bit · i-quant small — Importance-weighted 3-bit, small	Runs on 7× H100 SXM 80 GB
499.4 GB	502.7 GB	513.2 GB	459.7 GB	3-bit medium — 3-bit medium quantization	Runs on 7× H100 SXM 80 GB
550.5 GB	553.8 GB	564.4 GB	490.7 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Runs on 8× H100 SXM 80 GB
576.1 GB	579.4 GB	590 GB	501.3 GB	4-bit · i-quant non-linear — Importance-weighted 4-bit, non-linear	Runs on 5× H200 SXM 141 GB
576.1 GB	579.4 GB	590 GB	588.5 GB	4-bit small — 4-bit small quantization	Runs on 5× H200 SXM 141 GB
614.5 GB	617.8 GB	628.3 GB	629.6 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Runs on 5× H200 SXM 141 GB
704 GB	707.3 GB	717.9 GB	713.2 GB	5-bit small — 5-bit small quantization	Runs on 6× H200 SXM 141 GB
729.6 GB	732.9 GB	743.5 GB	758 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Runs on 6× H200 SXM 141 GB
844.7 GB	848 GB	858.6 GB	846.6 GB	6-bit, very good — 6-bit quantization, very good quality	Runs on 7× H200 SXM 141 GB
1023.8 GB	1027.1 GB	1037.6 GB	1087.6 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Runs on 6× B200 192 GB
2047.0 GB	2050.3 GB	2060.9 GB	2046.7 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Runs on 4 nodes of 8× H100 SXM

Selected Q4_K_M

4-bit medium quantization

616.0GBat 16K context

Weights: 614.0 GB
KV cache: 1.8 GB
Overhead: 0.3 GB

Server-class: ~616 GB at 16K context · about 5× H200 SXM 141 GB

Which Hardware Can Run MiMo V2.5 Pro?

Q4_K_M · 614.5 GB

MiMo V2.5 Pro (Q4_K_M) requires 614.5 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 799+ GB is recommended. Using the full 1049K context window can add up to 112.5 GB, bringing total usage to 727.0 GB. No consumer GPU has enough memory.

Needs server-class hardware

MiMo V2.5 Pro runs on a multi-GPU server · Q4_K_M needs ~616 GB

Hardware that runs this model
Hardware	Quant	Memory	Speed	Context	Hardware page
Mac 512 GB · Mac Studio (M3 Ultra, 512GB)	Q3_K_M	500.9 GB	~20 tok/s	64K ctx	Mac Studio →	Q3_K_M · 500.9 GB · ~20 tok/s · 64K ctx
8× H100 SXM 80 GB · 640 GB	IQ4_XS	552.1 GB	—	128K ctx	H100 SXM →	IQ4_XS · 552.1 GB · 128K ctx
5× H200 SXM 141 GB · 705 GB	Q4_K_M	616 GB	—	128K ctx	H200 SXM →	Q4_K_M · 616 GB · 128K ctx
4× B200 192 GB · 768 GB	Q4_K_M	616 GB	—	512K ctx	B200 →	Q4_K_M · 616 GB · 512K ctx

Also: NVIDIA DGX A100 640GB · NVIDIA DGX H100

≤ 640 GB tier →

By VRAM tier · typical GPUs and Macs at Q4_K_M

Consumer GPUs (up to 32 GB) and Macs up to 256 GB: too small for this model (needs ≥ 283.5 GB)8 GB tier →12 GB tier →16 GB tier →24 GB tier →Macs →

How good is MiMo V2.5 Pro?

The llmrun Score merges 6 public benchmark boards into a single figure, where 100 represents the average for open models.

Score vs VRAM in the ≤640 GB tier

MiMo V2.5 Pro
Other models
Best Score for its VRAM
x: VRAM (GB, log scale) · y: llmrun Score

#14 of 96 that fit ≤640 GB

Each dot is a model at the best quant that fits ≤640 GB. The line joins the models with the best Score for their VRAM; a dot below it is beaten by a model that needs no more memory.

Full ≤640 GB ranking →

Benchmarks

LMArena Text
LMArena · Oct 2026Arena
1467.3
#7/187 open
Arena
LMArena WebDev
LMArena · Oct 2026Arena
1481.3
#18/52 open
Arena
SciCode
Epoch AILab-reported
50.2%
#10/66 open
Lab-reported
ALE-Bench
Epoch AILab-reported
899.8
#14/43 open
Lab-reported
ProofBench
Epoch AILab-reported
22.0%
#11/26 open
Lab-reported
CritPt
Epoch AILab-reported
4.0%
#18/66 open
Lab-reported

All 8 MiMo V2.5 Pro benchmarks

In the ≤640 GB class

96 models fit ≤640 GB

Same Score, only models that fit ≤640 GB at their best quant: this model's peers by memory need.

#1
MiMo V2.6 Pro RLXiaomi
133.4
614.1 GB
#2
MiMo V2.6 Flash RLXiaomi
131.9
186.5 GB
#3
GLM 5.3Z.ai
130.1
453.8 GB
#4
DeepSeek V4.1 FlashDeepSeek
129.5
458.4 GB
#5
DeepSeek V4 Flash 0731DeepSeek
127.6
183.0 GB
#6
Muse Spark Open 816B A42BMeta
125.3
551.9 GB

Full ≤640 GB ranking →

Alternatives to MiMo V2.5 Pro

Official models with an llmrun Score, each answering one question; sizes are each model's default quant (usually Q4_K_M) with a 16K-token context.

Best alternative at this size
MiMo V2.6 Pro RL
Xiaomi
- 17 higher Score
- 2 GB less VRAM
Score 133.4 · 614.1 GB
Compare with MiMo V2.5 Pro Open MiMo V2.6 Pro RL
Fits in less VRAM
GLM 5.3
Z.ai
- 13.7 higher Score
- 162.2 GB less VRAM
Score 130.1 · 453.8 GB
Compare with MiMo V2.5 Pro Open GLM 5.3
Other vendor
DeepSeek V4.1 Flash
DeepSeek
- 13.1 higher Score
- 157.6 GB less VRAM
Score 129.5 · 458.4 GB
Compare with MiMo V2.5 Pro Open DeepSeek V4.1 Flash

Similar models

Model	Score	VRAM	tok/s	Why it is similar
DeepSeek V4 Pro 0813DeepSeek	129.2	990.8 GB	—needs > 24 GB	similar score
Muse Spark Open 816B A42BMeta	125.3	551.9 GB	—needs > 24 GB	similar score
GLM 5.2Z.ai	122.9	453.8 GB	—needs > 24 GB	similar score
Qwen3.8 27BAlibaba	122.1	17.9 GB	~38	similar score
Kimi K2.6Moonshot AI	120.9	617.6 GB	—needs > 24 GB	similar score
Inkling SmallThinking Machines	119.6	162.7 GB	—needs > 24 GB	similar score
GLM 5.1Z.ai	117.8	454.1 GB	—needs > 24 GB	similar score
DeepSeek V4 ProDeepSeek	116.5	959.8 GB	—needs > 24 GB	similar score
MiniMax M3MiniMax	112.5	258 GB	—needs > 24 GB	similar score
GLM 5Z.ai	111.6	454.1 GB	—needs > 24 GB	similar score

Specifications

Publisher: Xiaomi
Parameters: 1023.2B
Active parameters: 43.9B per token
Architecture: MiMoV2ForCausalLM
Context Length: 1,048,576 tokens
Vocabulary Size: 152,576
Release Date: 2026-04-27
License: MIT

Resources

Hugging Face model card (opens in a new tab)

Frequently Asked Questions

How much VRAM does MiMo V2.5 Pro need?

MiMo V2.5 Pro requires 614.5 GB of VRAM at Q4_K_M, or 2047.0 GB at BF16. Full 1049K context adds up to 112.5 GB (727.0 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 1023.2B × 4.8 bits ÷ 8 = 613.9 GB

KV Cache + Overhead ≈ 0.6 GB (at 2K context + ~0.3 GB framework)

Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.

KV Cache + Overhead ≈ 113.1 GB (at full 1049K context)

VRAM usage by quantization

Q4_K_M

614.5 GB

Q4_K_M + full context

727.0 GB

Learn more about VRAM estimation →

Can NVIDIA GeForce RTX 5090 run MiMo V2.5 Pro?

No — MiMo V2.5 Pro requires at least 281.9 GB at IQ2_XXS, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.

What's the best quantization for MiMo V2.5 Pro?

For MiMo V2.5 Pro, Q4_K_M (614.5 GB) offers the best balance of quality and VRAM usage. Q5_K_S (704 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 281.9 GB.

VRAM requirement by quantization

IQ2_XXS

281.9 GB~53%

Q2_K

435.4 GB~75%

IQ4_NL

576.1 GB~88%

Q4_K_M ★

614.5 GB~89%

Q5_K_S

704.0 GB~92%

BF16

2047.0 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run MiMo V2.5 Pro on a Mac?

Yes, but only at lower quantizations. The smallest Mac that can run MiMo V2.5 Pro is Mac Studio (M3 Ultra, 512GB) at Q3_K_M; 1 of the 146 Macs we list can run it at some quantization. For Q4_K_M (614.5 GB) you need a Mac with more unified memory.

Can I run MiMo V2.5 Pro locally?

Yes — MiMo V2.5 Pro can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

What's the download size of MiMo V2.5 Pro?

At Q4_K_M, the download is about 613.95 GB. The full-precision BF16 version is 2046.49 GB. The smallest option (IQ2_XXS) is 281.39 GB.

Which GPUs can run MiMo V2.5 Pro?

No single consumer GPU has enough VRAM to run MiMo V2.5 Pro at Q4_K_M (614.5 GB). Multi-GPU or professional hardware is required.

Which devices can run MiMo V2.5 Pro?

5 GB).