Tools & workflow · Case study
Fable 5.2 vs Opus 5.5: Which AI Makes Better Roblox Videos? (Blind Test)
A blind test where Opus 5.5 at max effort and Fable 5.2 at high effort each made the same Roblox-style animated cartoon in Claude Code from one shared script, voice track and studio. Each did a first draft, a critique, one fix round and a freeze; a blind judge compared 10 moments and checked adherence to the team's production rulebook. Tools include ElevenLabs, Remotion for the edit and the team's own Roblox-look kit. The description says the result is one test, not a benchmark. Summary from the description.
Disclosure Uses the channel's own engine and kit; promises a how-to follow-up.
Read at source