Six AI coding agents face off building 15-second videos from the same prompt
daradoescode · x · 2026-10-13
The post links to the Motion section of Which AI Made This: each model receives the same prompt inside its lab's own coding agent—make three different 15-second launch videos for a "second brain" note-taking app—building each as a web page in one session, then rendering all pages to video the same way.
Conditions are tightly controlled: fresh sessions with no skills, plugins, MCP servers, or custom instructions, so motion quality differences reflect the models themselves. Contenders include Opus 5.5 ($34.38 session cost), Fable 5.1 ($18.29), Haiku 5.5 ($5.13), GPT-6 Astra, GPT-6.1 Sol, and Muse Wonder 1.0 ($0.99). Hover a card to play the videos side by side.
More from Models
- Ex-exMVP: post-training should add a facet for how well models explain things — MParakhin · 2026-10-13
- GLM-5.3 abliterated for a month, no takeover: dev rails against AI doom narratives — mariofilhoml · 2026-10-13
- Claude auto-files a bug report about itself: TLDRs still too long after user asked for shorter — andrey_kurenkov · 2026-10-13
- BridgeBench one-prompt test shows Muse Wonder is fast and cheap but worst-looking — PVORY · 2026-10-13
- Meta skipped the 'too dangerous to release' marketing for Muse Wonder — and could rival Google — haider1 · 2026-10-13
- Anthropic says new defenses cut distillation attack data harvesting by 90-99%, cites Zhipu's GLM-5.3 — AnthropicAI · 2026-10-13