The creator's post

Why did OpenAI just make GPT-6.1 Sol twice as fast? Check out my test from yesterday. Opus 5.5 vs Sonnet 5.5 vs Sol 6.1 vs Grok 4.7 I gave each model 20 mins on medium to build a full stack app. I put urgency into the prompt which reminded each model every two minutes that time was running out. Look at how fast Opus and Sonnet finished compared to Sol. The work was all graded by a tester model and the app output is in the end of the video. I think Grok 4.7 might have the coolest UI.