Testing MiniMax H3 ref2v for Genshin Impact Anime Style Generation
ForsakenContract1135 · reddit · 2026-08-29
The author shares test results of generating Genshin Impact style anime videos using MiniMax H3's ref2v feature. By uploading reference images and combining them with prompts generated by Grok, they achieved action and scene generation. The workflow incorporates light2x 8 steps LoRA and seg att, with clips lasting 8-10 seconds each.
Related event: Hands-On: MiniMax H3 ref2v Generates Genshin-Style Anime Videos(2 posts)→
More from Multimodal
- Halcyon early-fusion image models match late-fusion with 275x less pretraining compute, open-sourced — davidwromero · 2026-10-16
- Running Wan2.1 14B FP16 on a 12GB laptop GPU: VRAM swap stress test finishes in 4m11s — RYuu-1120 · 2026-10-16
- Striking Midjourney generations: black holes and planetary aesthetics — ciguleva · 2026-10-16
- Fable 5.5 may drop today as Andrew Curran hints at early release — kimmonismus · 2026-10-16
- Kling 4.0 Adds Timeline Editing, Endorsed by 'Seoul Spring' Director Kim Seong-su — xiaohu · 2026-10-16
- One prompt, no reference images: Argon builds 3D Gaussian splats from geometry equations in NumPy — fofrAI · 2026-10-16