A full AI music video with MiniMax H3 Max Turbo: 24 scenes stitched by Python scripts
Inevitable_Emu2722 · reddit · 2026-09-04
A Redditor shares how they built a complete music video with MiniMax H3 Max Turbo instead of loose clips: timestamping the song into 24 scenes (mostly 6-10s), writing per-scene image and image-to-video prompts to keep characters and style consistent, generating via fal with a small Python script, then using another script to trim, order, strip generated audio and swap in the original MP3, finishing with scanlines in DaVinci Resolve. Not fully automatic—continuity mistakes like a phone facing the wrong way required retries—but H3 handled character motion surprisingly well.
More from Multimodal
- Voxtral 3 ships a LiveKit-based voice agent SDK streaming listen-think-speak — MistralAI · 2026-10-21
- Voxtral Mini TTS 3 hits SOTA expressiveness in 35+ languages with tag-based control — MistralAI · 2026-10-21
- Mistral launches Voxtral 3 speech models and new Voice Agents API — MistralAI · 2026-10-21
- Open-sourced pixel art normals generator ships as Rust CLI and wasm library — HugoDuprez · 2026-10-20
- Fibo 2.0 is strongest for ecommerce and creator content, weakest at UIUX design — ArtificialAnlys · 2026-10-20
- Fibo 2.0 leads old Fibo on all 9 capabilities, best at reasoning and anatomy — ArtificialAnlys · 2026-10-20