gemini-4POST

Gemini 4 Argon composition test: beats GPT but trails Opus 5.5

Given the same prompt to write code that generates a MIDI seaside-stroll song, Gemini 4 Argon beats both GPT versions but falls clearly short of Opus 5.5.

@ScarletKcArchived 20 days agoOriginally posted 2026.10.14

KC @ScarletKcOriginally posted 2026.10.14

Open original on X

Variora 的海边散步作曲题，Gemini 4 Argon 也交卷了同一份提示词，写一首动漫风的海边散步曲，模型要自己写代码生成 MIDI，再渲染成 MP3 Gemini 4 Argon 这首我听下来感觉一般般。比起 GPT 那两版是好了不少，但跟 Opus 5.5 放在一起差距太明显了，作曲这块 Opus 5.5 是真的无敌 Gemini 4 Argon 这版直接听 https://variora.fog.moe/zh/listen/?project=seaside-stroll&model=gemini-4-argon-high

x.com/scarletkc/status/2110058429964542178Archived 2026.10.14 10:24