I ran test between GLM 5.3, Opus 5 & GPT-5.6 Sol

3 models, Same prompts using /design feature from Command Code. Reviewed gameplay UX/UI and cost. - GLM 5.3: best pricing + solid UI - Opus 5: best Subway Surfers style clone - GPT‑5.6 Sol: nice UI, but the game isn’t playable

Naymur RahmanAug 143

GLM 5.3 made this song from scratch.

Much better than what I got from Opus 5.

⚡AI Search⚡Aug 1452

I asked Opus 5 (effort: high) to make its own representation of a total solar eclipse.

This is what came back. No textures, no models, no photographs, no ephemeris - about 2,300 lines of threejs and nothing else. Sea, sky, clouds, the lunar limb and the corona are all generated

CryptorAug 142

Claude opus 5 averaged 1.70km of travel a day.

Gpt-5 got 1.80, gpt-4o 2.57, and qwen2.5 3b running on my laptop got 4.15. clustering the same places by coordinates got it down to 0.80km. I compared against whichever model won each city, and still didn't lose one. 8 wins, 2

JC BuildsAug 142