AnalysisAI ModelsAugust 22, 2026

Stealth model Ox Alpha matches GPT-5.6 Sol mid on DeepSWE

Ox Alpha scored 96% on SWE-bench Verified Mini (48/50 tasks) and over 80% on 10 DeepSWE tasks, beating Fable (65%) and GPT-5.6-sol (52%). It performs on par with GPT-5.6 Sol mid, possibly a Chinese model like GLM-5.3 Flash.

How this story unfolded

2 days · 0 reports · 4 community posts · from Aug 21

  1. Aug 21
  2. Aug 22

Daily brief

Get tomorrow's AI brief in your inbox

More stories today

Open the live feed