MiniMax H3 ecosystem daily: 1920s horror LoRA, 8Mpx still-node tests, and new training adapter
optimisticalish · reddit · 2026-09-27
A roundup of September 26 additions to the MiniMax H3 image/video generation ecosystem:
- 1920s horror movies LoRA: trained on American clips, knows mid-1920s New York cityscapes, no trigger needed
- ComfyUI-Fizgig-H3-Still: nodes + workflows from the Fizgig maker; single-frame renders are now much faster than the old 5-frame workaround. Author's 8Mpx / 30-step test with Kitchen Attention: tiny lettering still fails, but film star names and studio credits in smaller type now work; it strips color casts (sepia becomes stark b&w) in exchange for smooth output with no pixel seams or scan lines
- Dark Life style LoRA: between woodcut/linocut and vintage b&w manga
- Head-mounted FPV camera LoRA for first-person videogame looks
- Ostris's H3 Training Adapter updated to v3
Links to two weeks of previous daily roundups are included.
More from Multimodal
- Comment: China leads the AI video revolution while the West argued about AI art — TinfoilTricorn · 2026-11-11
- "QT your Geisha" meme takes over X with stunning porcelain-style AI portraits — LudovicCreator · 2026-11-11
- MiniMax H3-Image early access test: 4 images, one coherent cyberpunk world — Aiden_Tech_Ai · 2026-11-11
- MiniMax H3 Image generates a GTA6-style hurricane scene from a single prompt — socialwithaayan · 2026-11-11
- Dragon Ball reimagined as a 1971 movie: another hit AI retro trailer — techhalla · 2026-11-11
- ReCaVSR lands in ComfyUI: one-step Wan2.2 video upscaler hits 4K at 4.65 fps on an RTX 4090 — LatentSpacer · 2026-11-11