Same 32 models benchmarked on four Macs: M5 Max matches M3 Ultra on MoE
Striking-Swim6702 · reddit · 2026-10-31
A Reddit user ran identical benchmarks across M2 Pro, M4 Pro, M5 Max, and M3 Ultra overnight with the same build, models, and settings. Key findings:
- MoE models favor GPU over bandwidth: with only 3-4B active params per token, the 36GB M5 Max matches or beats the 256GB M3 Ultra on Qwen3.5 4B (133), gpt-oss 20B (114), and Gemma 4 26B (115)
- Dense models remain bandwidth-bound: the Ultra still wins on 27B dense (37 vs 28)
- Image generation sees the biggest jump: M5 Max is 7x faster than M4 Pro and 2x faster than Ultra
- Text generation only improved 1.3-1.5x from M2 Pro to M4 Pro
More from Infra
- SenseNova-U1.5-8B hits 125 tok/s on M5 Ultra in llama.cpp and spots a $1 receipt error — Binary_orchid · 2026-10-31
- Cloudflare Durable Objects now defer code updates up to 300s by default — irvinebroque · 2026-10-31
- Intel plunges 29%, falling below its 2000 dot-com era peak after 26 years — mitchdeg · 2026-10-31
- Clover-35B-A3B Quantized to 29.4 GB Runs at ~42 tok/s on a 64 GB Mac — viveky259 · 2026-10-31
- Liquid Agent brings fully local LFM3 agent to Linux with one-command install — JosephJacks_ · 2026-10-31
- A fully local office AI stack: MiniCPM, Qwen3, LFM2.5 and Whisper — HowDevelop · 2026-10-31