MacBook Pro M4 ComfyUI test: 3-min images, but a 2-sec video takes 9+ hours
Reaperman-13 · reddit · 2026-09-13
A Reddit user shares real-world results running generative models locally in ComfyUI on a MacBook Pro M4:
- Text-to-image: Z Image Turbo works at 1024x1024 in 3-4 minutes (workflow-dependent, 2.5-12 min range); Krea takes 6 minutes.
- Music: MiniMax Music was the only model that ran out of the box, but a single generation ran 12 hours before being cancelled.
- Video: MiniMax video needed 9+ hours for a 2-second 0.2MP clip — effectively unusable.
- The retired author living in Asia can't build a Windows rig; the Mac is fine for video editing and upscaling, but local AI generation is clearly not its strength.
More from Infra
- vLLM Dev Pushes Back as 5 Specialized Inference Engines Launch in One Month — AccBalanced · 2026-09-13
- Ex-compute veteran rebuts Dario's 'pace the frontier': compute doesn't idle, it reroutes — basedjensen · 2026-09-13
- Pirate Face mirrors 669k+ Hugging Face models as checksum-verified torrents — cephaloform · 2026-09-13
- Cloud agents beat local Mac scaling, argues altryne amid parallel agent experiments — altryne · 2026-09-13
- UBTech's 10,000-Unit Humanoid Robot Factory Starts Production, One Robot Every 10 Minutes — 创业邦 · 2026-09-13
- Sell an RTX 5090 for a Mac Studio M5 Ultra 96GB? Weighing 1.8 vs 1.2 TB/s Bandwidth for Local AI — unchikuso · 2026-09-13