Redditor struggles with outfit swap: Qwen Image 2.1 breaks body proportions
Existing-Box-7866 · reddit · 2026-10-27
A Reddit user is asking for help building an outfit-swap workflow for AI characters: take multi-angle reference images of a character and re-dress them in a new outfit for later image/video generation.
Their findings so far:
- Qwen Image 2.1 (i2i): keeps head, face and hair nearly identical, but consistently breaks body proportions (bust, shoulders, arms). They tried many workflows, custom nodes, LoRAs (e.g. outfit swapper), prompt strategies, samplers and multi-refs — none worked.
- Flux 3 Klein: preserves proportions better but not perfectly, with worse overall quality and coherence.
- GPT Image: strong results but frequent safety refusals even for non-NSFW requests.
No solid solution has emerged yet; worth watching the thread for follow-ups.
More from Multimodal
- Krea for Premiere Pro confirmed on official site, AI generation in timeline — lmoroney · 2026-10-27
- Krea ships Premiere Pro plugin to generate video and SFX inside the timeline — lmoroney · 2026-10-27
- Gemini generates an illustrated morning mobility cheat sheet from YouTube shorts — FanaHOVA · 2026-10-27
- Wan2GP lands on Pinokio: 1-click AI video on 6GB GPUs — cocktailpeanut · 2026-10-27
- WanGP v18.81 adds Tencent's Hy-Nomas world model, playable on 8GB VRAM — cocktailpeanut · 2026-10-27
- Butter.video wins over skeptical creator with hundreds of tasteful timeline 'blocks' for video effects — beechinour · 2026-10-27