We introduce FireRed-Image-Edit-1.1, an upgrade to our general-purpose image editing foundation model. Building upon the capabilities presented in the FireRed-Image-Edit-1.0 Technical Report, version 1.1 significantly enhances identity consistency, multi-image conditioning, and domain-specialized editing performance, bringing the model closer to real-world creative production needs.
✨ What's New
Stronger Identity Consistency — Substantially improved facial and character consistency during edits, with better texture stability and reduced identity drift in portrait and cosmetic scenarios.
Enhanced Multi-Image Injection — Native support for multi-image conditioning enables more reliable identity transfer, reference-guided editing, virtual try-on, and group photo composition.
Artistic Typography (CoverCraft) — Improved decorative font rendering and stylized text generation for cover design, poster creation, and expressive title layouts while preserving structural clarity.
Professional Makeup Editing (Makeup LoRA) — Refined cosmetic editing capabilities including skin enhancement, color harmonization, and detailed eye and lip adjustments, with strong realism and identity preservation.
High-Speed Distilled Variant (Lightning 8-step) — An optimized 8-step distilled model that significantly reduces inference latency while maintaining strong instruction alignment and visual coherence.
