local.ailocal.aiEarly access

fuxiaopang

local.ai/fuxiaopang

Claimed August 2026

My machines

Mac Studio M4 Max
Approximate
Apple · 36 GB
Approximate options
Choose what matters most
Best fit
Gemma 4 26B A4B It · 4BIT
Best balance of speed and intelligence.
Intelligence57.1
Speed20.1 tasks/hr
Size15.3 GB
More intelligent
Qwen3.6 35B A3B · UD-Q3_K_XL
~23% more intelligence · ~51% slower.
Intelligence70.4
Speed13.3 tasks/hr
Size16.8 GB
Faster
Gemma 4 E4B It · Q4_K_M
~70% faster · retains ~50% intelligence.
Intelligence28.3
Speed67.3 tasks/hr
Size5.0 GB
Approximate for 36 GB VRAM. Results may vary.