Every memory option compared, 150 GB/s memory bandwidth. Updated 2 November 2026.
M3 Pro is sold with 18, 36 GB of memory. Counting the 31 models tracked here: 18 GB runs 16 models well; 36 GB runs 23 models well. Memory bandwidth is 150 GB/s, about 1.3 times the memory speed of a base M4 (120 GB/s), so the same model generates words proportionally faster. Apple Silicon shares one pool of memory between the processor and the GPU, and macOS lets the GPU use about two thirds of it on Macs up to 36 GB and about three quarters above that, so the memory you buy decides which models load at all.
| Model | 18 GB | 36 GB |
|---|---|---|
| Wan 2.2 14B | No | Slow about 53 min |
| LTX-2.3 | No | Good about 4 min |
| Qwen 3.6 35B A3B | No | Tight 20/s |
| Qwen 3.5 35B A3B | No | Great 23/s |
| Gemma 4 31B | No | Slow 4/s |
| DeepSeek R1 32B | No | Slow 4/s |
| Qwen3 Coder 30B | No | Great 22/s |
| Qwen 3.6 27B | No | Slow 5/s |
| Gemma 4 26B A4B | No | Great 17/s |
| Z-Image Turbo 6B | Slow about 2 min | Good about 49 s |
| Wan 2.2 5B | Slow about 27 min | Good about 13 min |
| gpt-oss 20B | Tight 16/s | Great 18/s |
| ACE-Step 1.5 | Good about 3 min | Good about 3 min |
| FLUX.2 klein 4B | Good about 34 s | Good about 34 s |
| Phi-4 14B | Good 9/s | Good 9/s |
| Gemma 4 12B | Good 10/s | Good 10/s |
| Stable Diffusion XL | Good about 18 s | Good about 18 s |
| Qwen3 TTS 1.7B | Good about 52 s | Good about 52 s |
| Gemma 4 E4B | Great 25/s | Great 25/s |
| Qwen 3.5 9B | Great 12/s | Great 12/s |
| Qwen3 VL 8B | Great 13/s | Great 13/s |
| DeepSeek R1 8B | Great 15/s | Great 15/s |
| Qwen2.5 Coder 7B | Great 17/s | Great 17/s |
| Gemma 4 E2B | Great 40/s | Great 40/s |
| Whisper large-v3 turbo | Good about 1 min | Good about 1 min |
| Qwen 3.5 4B | Great 24/s | Great 24/s |
| Kokoro TTS | Good about 11 s | Good about 11 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 17 GB is usable by the GPU. The largest chat or coding model that fits is gpt-oss 20B (15.2 GB), generating about 16 words a second with 1.9 GB to spare. The quickest is Gemma 4 E2B at about 40 words a second. Closest miss: Gemma 4 26B A4B, which needs about 19.4 GB (2 GB more than this configuration can give the GPU).
Find M3 Pro Macs with 18 GB on Amazon
About 34 GB is usable by the GPU. The largest chat or coding model that fits is Qwen 3.6 35B A3B (24.6 GB), generating about 20 words a second with 9.6 GB to spare. The quickest is Gemma 4 E2B at about 40 words a second. Closest miss: Llama 3.3 70B, which needs about 45.6 GB (11 GB more than this configuration can give the GPU).
For everyday chat, 18 GB runs 16 of 31 models well. 18 GB to 36 GB unlocks 7 more: Gemma 4 26B A4B, Qwen 3.5 35B A3B, Qwen3 Coder 30B, Qwen 3.6 35B A3B, Z-Image Turbo 6B and more.
No. It needs about 21.5 GB of usable memory and this chip's largest option is 36 GB.
18 GB runs gpt-oss 20B at about 16 words a second; 36 GB runs gpt-oss 20B at about 18 words a second.
As an Amazon Associate I earn from qualifying purchases.
Unlocks 4 more: Gemma 4 31B, Qwen 3.6 27B, DeepSeek R1 32B, Wan 2.2 14B.
Unlocks 7 more: Gemma 4 31B, Llama 3.3 70B, Qwen 3.5 122B A10B, gpt-oss 120B and more.