Moss image model natively generates consistent same-character comics in any style
AltruisticList6000 · reddit · 2026-10-24
A Reddit user showcases the Moss SFT int8 image model (merged with a stripe fix LoRA and Qwen 9b int8) generating a four-panel Hatsune Miku comic with the same character kept consistent across panels and switchable styles including photorealism, all natively at 1080p.
Highlights:
- Extremely strong prompt adherence with simple handwritten sentences — no prompt enhancer, JSON, or bbox needed
- Experimental turbo method: dmd2 turbo LoRA with high CFG and negative prompts to fix blur and bokeh, still quite fast
- Author sees big potential for image editing and editing LoRAs
More from Multimodal
- Striking MidJourney --sref demo turns Superman into motion-blurred abstract art — gizakdag · 2026-10-24
- Early Kimi K3.1 Pro access: $4 in tokens, 3-4 hours for Blender clay renders — ChrisGPT · 2026-10-24
- Video DiT runs on 12GB VRAM at same speed via int8, model weighs only 9GB — cocktailpeanut · 2026-10-24
- MOSS review: best prompt comprehension since SDXL era, native 2MP output — Perfect-Campaign9551 · 2026-10-24
- RTX 30-series GPU Unlock tested: up to 59% throughput gain in ComfyUI on 3090 Ti — under-_-null · 2026-10-24
- Blender + MiniMax H3 workflow recreates a Spider-Man swing through a 3D city — Fun_Cranberry4894 · 2026-10-24