Every memory option compared, 256 GB/s memory bandwidth. Laptops with this chip include ASUS ROG Flow Z13, HP ZBook Ultra G1a. Updated 2 November 2026.
Ryzen AI Max+ 395 is sold with 32, 64, 128 GB of memory. Counting the 31 models tracked here: 32 GB runs 23 models well; 64 GB runs 25 models well; 128 GB runs 27 models well. Memory bandwidth is 256 GB/s, about 3.2 times the memory speed of an everyday laptop without an AI chip (80 GB/s), so the same model generates words proportionally faster. These laptops share their RAM between the processor and the built-in graphics, and the RAM is soldered, so it cannot be upgraded later. Windows also needs its share, so roughly 70% of the RAM is realistically available for a model, and the memory you buy decides which models load at all.
| Model | 32 GB | 64 GB | 128 GB |
|---|---|---|---|
| Qwen 3.5 122B A10B | No | No | Good 11/s |
| gpt-oss 120B | No | No | Great 21/s |
| Llama 3.3 70B | No | Slow 3/s | Slow 3/s |
| Wan 2.2 14B | Slow about 28 min | Great about 14 min | Great about 14 min |
| LTX-2.3 | Great about 2 min | Great about 2 min | Great about 2 min |
| Qwen 3.6 35B A3B | No | Great 33/s | Great 33/s |
| Qwen 3.5 35B A3B | Great 35/s | Great 35/s | Great 35/s |
| Gemma 4 31B | Slow 6/s | Slow 6/s | Slow 6/s |
| DeepSeek R1 32B | Slow 6/s | Slow 6/s | Slow 6/s |
| Qwen3 Coder 30B | Great 32/s | Great 32/s | Great 32/s |
| Qwen 3.6 27B | Good 6/s | Good 6/s | Good 6/s |
| Gemma 4 26B A4B | Great 25/s | Great 25/s | Great 25/s |
| Z-Image Turbo 6B | Great about 27 s | Great about 27 s | Great about 27 s |
| Wan 2.2 5B | Great about 7 min | Great about 7 min | Great about 7 min |
| gpt-oss 20B | Great 26/s | Great 26/s | Great 26/s |
| ACE-Step 1.5 | Great about 1 min | Great about 1 min | Great about 1 min |
| FLUX.2 klein 4B | Great about 19 s | Great about 19 s | Great about 19 s |
| Phi-4 14B | Great 13/s | Great 13/s | Great 13/s |
| Stable Diffusion XL | Great about 11 s | Great about 11 s | Great about 11 s |
| Qwen3 TTS 1.7B | Great about 30 s | Great about 30 s | Great about 30 s |
| Gemma 4 12B | Great 15/s | Great 15/s | Great 15/s |
| Gemma 4 E4B | Great 36/s | Great 36/s | Great 36/s |
| Qwen 3.5 9B | Great 17/s | Great 17/s | Great 17/s |
| Qwen3 VL 8B | Great 19/s | Great 19/s | Great 19/s |
| DeepSeek R1 8B | Great 23/s | Great 23/s | Great 23/s |
| Qwen2.5 Coder 7B | Great 25/s | Great 25/s | Great 25/s |
| Gemma 4 E2B | Great 58/s | Great 58/s | Great 58/s |
| Whisper large-v3 turbo | Great about 42 s | Great about 42 s | Great about 42 s |
| Qwen 3.5 4B | Great 35/s | Great 35/s | Great 35/s |
| Kokoro TTS | Great about 6 s | Great about 6 s | Great about 6 s |
Text models show words a second. Image, video and voice models show the time per job (image, 4 second clip, minute of speech or similar).
As an Amazon Associate I earn from qualifying purchases.
About 24 GB is usable for models. The largest chat or coding model that fits is Qwen 3.5 35B A3B (23.6 GB), generating about 35 words a second with 0.4 GB to spare. The quickest is Gemma 4 E2B at about 58 words a second. Closest miss: Qwen 3.6 35B A3B, which needs about 24.6 GB (1 GB more than this configuration can give the GPU).
Find Ryzen AI Max+ 395 laptops with 32 GB on Amazon
About 48 GB is usable for models. The largest chat or coding model that fits is Qwen 3.6 35B A3B (24.6 GB), generating about 33 words a second with 23.4 GB to spare. The quickest is Gemma 4 E2B at about 58 words a second. Closest miss: gpt-oss 120B, which needs about 68.8 GB (21 GB more than this configuration can give the GPU).
Find Ryzen AI Max+ 395 laptops with 64 GB on Amazon
About 96 GB is usable for models. The largest chat or coding model that fits is Qwen 3.5 122B A10B (85.5 GB), generating about 11 words a second with 10.5 GB to spare. The quickest is Gemma 4 E2B at about 58 words a second. Closest miss: DeepSeek V3 671B, which needs about 424.7 GB (329 GB more than this configuration can give the GPU).
For everyday chat, 32 GB runs 23 of 31 models well. 64 GB to 128 GB unlocks 2 more: Qwen 3.5 122B A10B, gpt-oss 120B.
No. It needs about 21.5 GB of usable memory and this chip's largest option is 128 GB.
32 GB runs gpt-oss 20B at about 26 words a second; 64 GB runs gpt-oss 20B at about 26 words a second; 128 GB runs gpt-oss 20B at about 26 words a second.
As an Amazon Associate I earn from qualifying purchases.
Unlocks 2 more: Gemma 4 31B, DeepSeek R1 32B.