A creator runs Grok 4.6 through many front-end and 3D coding tests, calling it close to but not quite at Sonnet/GPT-tier quality.
Continue to AI University →