Lyria 3.5 called the best music model yet as songs sound more real
AndyMasley · x · 2026-09-06
AndyMasley says Lyria 3.5 is the best music model right now, with generated songs sounding a little more real each iteration, sharing samples. A counterpoint to claims that AI music composition still falls short.
Related event: Google's Lyria 3.5 Praised as Best Music Generation Model(2 posts)→
More from Multimodal
- Tencent's Hy4 video model is quietly popping off, says AI commenter — zainhas · 2026-09-06
- Full Seedance 2.5 prompt: single-take paparazzi-style AI video, copy-paste ready — techhalla · 2026-09-06
- A Seedance 2.5 prompt that makes AI video look like raw paparazzi footage — techhalla · 2026-09-06
- 4-step Minimax H3 video generation lands: 3080ti render time drops from 54 to 12 minutes — cocktailpeanut · 2026-09-06
- Emad Mostaque: Don't ask Astra to make biblically accurate angels — PeterBowdenLive · 2026-09-06
- Precise camera control and bullet time with Minimax-H3, single image input — init-5 · 2026-09-06