M3 Ultra Local LLM Tests Show Coding Still Out of Reach

Tests on a 512GB M3 Ultra show GLM5.2 running at about 16 t/s, fine for background tasks like media organization but too slow for heavy daily coding workflows.

2026-08-26 ~ 2026-08-26 · 2 related posts

• M3 Ultra local benchmark: GLM5.2 and Qwen speeds limited — natesiggard · 2026-08-26
• Local inference struggles with coding at 16 t/s, despite media use cases — natesiggard · 2026-08-26