Skip to content

What can the Ryzen AI Max+ 395 64 GB run?

What fits on the Ryzen AI Max+ 395 64 GB
Text models: 181Visual models: 20Strongest first
Memory
64 GB unified
Bandwidth
256 GB/s
Maker
AMD
Also called
Strix Halo

Open the calculator on this machineRyzen AI Max+ 395 (Strix Halo) in our inventory

Text models that run on it

Text models that run on the Ryzen AI Max+ 395 64 GB, best first
ModelBest fitContextSpeedStrongest at
Qwen3.8 27B8-bit256K5.3 tok/sChat, Coding, Agents, Docs
Muse Glimmer 30B8-bit128K5.0 tok/sDocs
Qwen3.5 122B A10B2-bit40K34 tok/sChat, Coding, Agents, Docs
Nemotron 3 Super 120B A12B2-bit208K26 tok/sChat, Coding, Agents, Docs
Nemotron 3.5 Bolt 74B A9B4-bit256K22 tok/sChat, Coding, Agents, Docs
Qwen3.6 35B A3B8-bit256K40 tok/sChat, Coding, Agents, Docs
Gemma 4 31B8-bit256K4.6 tok/sChat, Coding, Agents, Docs
Gemma 4 26B A4B8-bit256K34 tok/sChat, Coding, Agents, Docs
GLM-4.7 Flash8-bit198K40 tok/sChat, Coding, Agents
Gemma 4 12B8-bit256K12 tok/sChat, Coding, Docs
Mistral Small 4 119B2-bit144K52 tok/sChat, Docs
gpt-oss 120B2-bit112K64 tok/sChat, Coding
Qwen3 Coder Next 80B3-bit256K70 tok/sCoding, Agents
Nemotron 3.5 Lightning 30B A3B8-bit256K39 tok/sAgents, Docs
Devstral Small 2 24BFP8128K6.1 tok/sCoding
LFM2.5 24B A2B8-bit128K57 tok/sAgents
Mistral Small 3.2 24B8-bit128K6.1 tok/sChat, Coding, Agents
Ministral 3 14B8-bit192K11 tok/sCoding, Agents, Docs
gpt-oss 20BMXFP4128K53 tok/sAgents
Ministral 3 8B8-bit256K17 tok/sDocs
Llama 3.3 70B4-bit16K3.6 tok/sChat, Coding, Agents, Docs
Olmo 3.1 32B8-bit64K4.5 tok/sChat, Coding, Agents, Docs
Llama 3.1 8B8-bit128K19 tok/sChat, Coding, Agents, Docs
Qwen3 14B8-bit40K10 tok/sNot rated
Qwen3 VL 8B Instruct8-bit256K17 tok/sText and images
Qwen3 8B8-bit40K18 tok/sNot rated
Qwen3 32B8-bit40K4.5 tok/sNot rated
Qwen3.6 27B8-bit256K5.3 tok/sText and images
Qwen2.5 Coder 14B Instruct8-bit32K10 tok/sCoding
Qwen3 Coder 30B A3B Instruct8-bit152K40 tok/sCoding
Qwen3.5 27B8-bit256K5.3 tok/sText and images
Qwen3.5 35B A3B8-bit256K40 tok/sText and images
NVIDIA Nemotron 3 Nano 30B A3B BF168-bit256K39 tok/sNot rated
GLM 4.5 Air2-bit32K25 tok/sNot rated
Qwen3 30B A3B8-bit40K40 tok/sNot rated
Qwen3 30B A3B Instruct 25078-bit152K40 tok/sNot rated
Laguna XS.28-bit256K41 tok/sNot rated
OTel 2.0 LLM 31B IT8-bit256K4.6 tok/sNot rated
Qwen2.5 32B Instruct8-bit32K4.5 tok/sNot rated
MiniCPM V 4.7 35B A3B8-bit256K40 tok/sText and images
Qwen2.5 14B Instruct8-bit32K10 tok/sNot rated
Laguna XS 2.18-bit256K41 tok/sNot rated
Diffusiongemma 26B A4B IT8-bit256K33 tok/sText and images
Qwen2.5 VL 32B Instruct8-bit44K4.4 tok/sText and images
Qwen2.5 Coder 32B Instruct8-bit32K4.5 tok/sCoding
Laguna S 2.12-bit80K40 tok/sNot rated
Qwen3 VL 30B A3B Instruct8-bit144K40 tok/sText and images
EXAONE 3.5 7.8B Instruct8-bit32K19 tok/sNot rated
DeepSeek R1 0528 Qwen3 8B8-bit128K18 tok/sNot rated
Gemma 3 12B IT8-bit128K13 tok/sText and images
Phi 3.5 Mini Instruct8-bit104K37 tok/sNot rated
DeepSeek R1 Distill Qwen 32B8-bit48K4.5 tok/sNot rated
Gemma 3 27B IT8-bit128K5.5 tok/sText and images
Ornith 1.5 35B A3B8-bit256K40 tok/sNot rated
Ornith 1.5 122B2-bit40K27 tok/sNot rated
GPT Neox 20B8-bit2K7.0 tok/sNot rated
Apertus 8B Instruct 25098-bit64K19 tok/sNot rated
Phi 48-bit16K10 tok/sNot rated
Phi 3 Mini 128K Instruct8-bit104K37 tok/sNot rated
Qwen3 VL 32B Instruct8-bit44K4.4 tok/sText and images

Best fit is the highest quality, up to 8-bit, that leaves the use case’s context. Ranked by our use-case scores, then by Hugging Face downloads. Speeds are rough, plus or minus 25%. The top 60 of 181 are shown.

Image and video models

Image and video models that run on the Ryzen AI Max+ 395 64 GB
ModelBest fitPer outputStrongest at
MiniMax H34-bit32 minVideo
LTX-2.58-bit3.1 minVideo
Z-Image Turbo16-bit27 sImages
FLUX.1 [dev]16-bit53 sImages
Qwen-Image8-bit17 minImages
FLUX.2 [dev]5-bit8.4 minImages
Qwen-Image 2.116-bit7.1 minImages
FLUX.3 Klein16-bit39 sImages
Wan2.2 I2V A14B16-bit2.4 hVideo
FLUX.2 [klein] 9B16-bit13 sImages
Wan2.1 I2V 14B16-bit2.4 hVideo
Wan 2.1 T2V-14B16-bit3.1 hVideo
Krea 2 Turbo16-bit52 sImages
Stable Diffusion 3.5 Large16-bit3.7 minImages
Qwen Image 25128-bit17 minImages
MOSS-Image16-bit4.6 minImages
Wan 2.2 TI2V-5B16-bit40 minVideo
MOSS Image Base16-bit4.6 minImages
Wan 2.2 A14B16-bit2.4 hVideo
HunyuanVideo 1.516-bit3.0 hVideo

Only just too big

These need up to a quarter more memory than the Ryzen AI Max+ 395 64 GB has. Spilling into system RAM or a lower precision can get them running, slowly.

Questions

What is the best LLM for the Ryzen AI Max+ 395 64 GB?

Of the open models we track, the strongest that run well on the Ryzen AI Max+ 395 64 GB are Qwen3.8 27B, Muse Glimmer 30B, Qwen3.5 122B A10B. 181 text models fit it in all.

Calculated from published specs, not benchmarked. All machines