Mac 128 GB vs RTX 5090 32 GB for local AI
The Mac 128 GB has 83.4 GB more memory, and that decides it for local AI: it runs a better file (or runs at all) for 22 of 49 models, and 49 run at 8-bit or better against 47. The RTX 5090 32 GB does have hardware FP8, so for models that fit on both it can be faster per step. Memory is only half the story here: the Mac holds more, but it is several times slower per image than a desktop RTX card, and FP8 files do not work on it.
Mac 128 GB
- VRAM
- 115.4 GB
- Memory
- Unified
- Bandwidth
- —
- Architecture
- Apple Silicon
- FP8 compute
- No
- Power
- —
VS
RTX 5090 32 GB
- VRAM
- 32 GB
- Memory
- GDDR7
- Bandwidth
- 1792 GB/s
- Architecture
- Blackwell
- FP8 compute
- Yes
- Launch price
- $1,999
01Model by model
| Model | Mac 128 GB | RTX 5090 32 GB | ||
|---|---|---|---|---|
| image models | ||||
| FLUX.1 dev | Runs well 16-bit | Runs well 16-bit | ||
| FLUX.1 schnell | Runs well 16-bit | Runs well 16-bit | ||
| FLUX.1 Kontext | Runs well 16-bit | Runs well 16-bit | ||
| FLUX.1 Krea | Runs well 16-bit | Runs well 16-bit | ||
| FLUX.1 Fill | Runs well 16-bit | Runs well 16-bit | ||
| FLUX.2 dev | Runs well 16-bit | Runs Q6_K | ||
| FLUX.2 klein 9B | Runs well 16-bit | Runs well 16-bit | ||
| FLUX.2 klein 4B | Runs well 16-bit | Runs well 16-bit | ||
| Krea 2 | Runs well 16-bit | Runs well 16-bit | ||
| Qwen-Image | Runs well 16-bit | Runs well FP8 | ||
| Qwen-Image-Edit | Runs well 16-bit | Runs well FP8 | ||
| Qwen-Image 2.1 | Runs well 16-bit | Runs well 16-bit | ||
| Z-Image Turbo | Runs well 16-bit | Runs well 16-bit | ||
| Z-Image | Runs well 16-bit | Runs well 16-bit | ||
| Ideogram 4 | Runs well Q8_0 | Runs well FP8 | ||
| Boogu-Image | Runs well 16-bit | Runs well 16-bit | ||
| ERNIE-Image | Runs well 16-bit | Runs well 16-bit | ||
| HiDream-O1 | Runs well 16-bit | Runs well 16-bit | ||
| Mage-Flow | Runs well 16-bit | Runs well 16-bit | ||
| Lumina 2.0 | Runs well 16-bit | Runs well 16-bit | ||
| HiDream-I1 Full | Runs well 16-bit | Runs well FP8 | ||
| HiDream-I1 | Runs well 16-bit | Runs well FP8 | ||
| SD 3.5 Large | Runs well 16-bit | Runs well 16-bit | ||
| SD 3.5 Medium | Runs well 16-bit | Runs well 16-bit | ||
| Chroma1-HD | Runs well 16-bit | Runs well 16-bit | ||
| SDXL | Runs well 16-bit | Runs well 16-bit | ||
| Illustrious / Pony | Runs well 16-bit | Runs well 16-bit | ||
| SD 1.5 | Runs well 16-bit | Runs well 16-bit | ||
| HunyuanImage 2.1 | Runs well 16-bit | Runs well FP8 | ||
| video models | ||||
| Wan 2.1 14B | Runs well 16-bit | Runs well FP8 | ||
| Wan 2.1 1.3B | Runs well 16-bit | Runs well 16-bit | ||
| Wan 2.1 I2V 480P | Runs well 16-bit | Runs well FP8 | ||
| Wan 2.1 I2V 720P | Runs well 16-bit | Runs well FP8 | ||
| Wan VACE 14B | Runs well 16-bit | Runs well Q8_0 | ||
| Wan 2.2 T2V | Runs well 16-bit | Runs well FP8 | ||
| Wan 2.2 I2V | Runs well 16-bit | Runs well FP8 | ||
| Wan 2.2 5B | Runs well 16-bit | Runs well 16-bit | ||
| Wan 2.2 Animate | Runs well 16-bit | Runs well FP8 | ||
| Wan Animate 2 | Runs well 16-bit | Runs well Q8_0 | ||
| Wan 2.2 S2V | Runs well 16-bit | Runs well FP8 | ||
| SCAIL-2 | Runs well 16-bit | Runs well FP8 | ||
| HunyuanVideo 13B | Runs well 16-bit | Runs well 16-bit | ||
| LTX-Video 13B | Runs well 16-bit | Runs well FP8 | ||
| HunyuanVideo 1.5 | Runs well 16-bit | Runs well 16-bit | ||
| LTX-2 | Runs well 16-bit | Runs well Q8_0 | ||
| LTX-2.3 | Runs well 16-bit | Runs well Q8_0 | ||
| LTX-2.5 | Runs well 16-bit | Runs well Q8_0 | ||
| MiniMax H3 | Runs well 16-bit | Runs Q5_K_M | ||
| MiniMax H3 Pruned | Runs well 16-bit | Runs well FP8 | ||
Highlighted: the GPU that runs a better file for that model. Launch prices are the maker’s original list prices, not today’s street prices. Calculated from real file sizes; how the numbers work.
02Score
49run well on Mac 128 GB
47run well on RTX 5090 32 GB
22better on Mac 128 GB
0better on RTX 5090 32 GB