Model Drop | ZIT + LTX 2.3 + Music Video | Arca Gidan contest
An AI artist details a complete workflow from lyrics to video using Suno AI, Z Image Turbo, and LTX 2.3.
An AI artist has gone viral by detailing their complete, end-to-end pipeline for creating a music video titled 'Model Drop' using only AI tools. The project began with original lyrics about the relentless pace of new AI model releases, which were then turned into a full song track using Suno AI. This audio track became the creative baseline for the entire visual project.
For the visuals, the creator developed a detailed shot list mapped to the song's structure. They generated a consistent main character using the Z Image Turbo (ZIT) model, achieving character continuity not through training a LoRA but through meticulous, identical prompting across dozens of scenes. The still images were then animated using the LTX 2.3 image-to-video model within ComfyUI, with specific notes on using the official workflow for best results and keeping shots medium-to-close for effective lip-sync. The final video was assembled using a mobile editing app, demonstrating a sophisticated, multi-stage creative process entirely powered by current generative AI tools.
- Used Suno AI to generate the complete song track from original lyrics about AI model FOMO.
- Created a consistent character across dozens of shots using Z Image Turbo (ZIT) with prompting, not a trained LoRA.
- Animated the stills with LTX 2.3's img2video, noting the official workflow is superior for lip-sync quality.
Why It Matters
It demonstrates a professional, integrated pipeline for AI-native content creation, moving beyond single-tool experiments.