Moss SFT tested on non-portrait scenes: photoreal car showroom generations
Perfect-Campaign9551 · reddit · 2026-10-19
A Reddit user shared images generated with the Moss SFT model in ComfyUI using the Layout workflow, focusing on non-generic "person standing / portrait" scenes to test styles and camera behavior.
The featured image — a vintage red Ford pickup in a showroom — shows highly photorealistic detail, lighting, and materials, highlighting the model's strength in object/scene photography rather than typical portrait output.
More from Multimodal
- French #slopcore AI track went viral without lyrics, now gets an English-subtitled version — banteg · 2026-10-19
- Krea2 Turbo 2MP Upscaling Fails Across Samplers, User Asks Reddit for a Working Workflow — AkaToraX · 2026-10-19
- GPT-6 Image 2.5 renders painterly 'The Cello Man' in striking detail — DeryaTR_ · 2026-10-19
- AI-generated parody ad for fictional "Curry King" men's fragrance amuses Reddit — IntelligentAirport36 · 2026-10-19
- Opus 5.5 builds a realistic 3D ocean exploration scene in 4,100 lines of code — techartist_ · 2026-10-19
- Cat DJ video generated on a handheld gaming PC: 3 seconds took 16 minutes — Top-Seaworthiness522 · 2026-10-19