ai.models.index.hero.badge

ai.models.index.hero.title

ai.models.index.hero.description

ai.models.index.sections.image_models

ai.models.index.sections.related_image_models

ai.models.index.sections.video_models

ai.models.index.sections.related_video_models

FAQ

Frequently Asked Questions

Common questions about FireRed Image Edit

A general-purpose image editing model by Xiaohongshu's Intelligent Creation Core Technology Team, built on Diffusion Transformer architecture with Qwen2.5-VL as vision-language encoder.

Yes. FireRed Edit and FireRed-Image-Edit both refer to the FireRed Image Edit model and its editing workflow. Some searches misspell the name as firerd image edit; the correct project name is FireRed Image Edit.

10+ categories: object add/remove/replace, attribute adjustment, background editing, style transfer, text editing, photo restoration, multi-image editing, virtual try-on, portrait makeup, and multi-element fusion.

30GB VRAM with optimized inference (distillation + quantization + static compilation), ~4.5s per sample.

Open-source SOTA on ImgEdit (4.56), GEdit EN (7.943), GEdit CN (7.887), REDEdit EN (4.26), REDEdit CN (4.33), surpassing some proprietary models.

Yes, native bilingual support for both Chinese and English editing instructions.

Automatic multi-image processing: ROI detection โ†’ crop & stitch โ†’ recaption. Supports 1-3 native input images, and 3+ via Agent.

Yes, full LoRA training code is released. Also provides LoRA Zoo with pre-trained styles (Makeup, Covercraft text style, etc.)

Apache 2.0, fully open source. Available on HuggingFace, ModelScope, and GitHub.

Testimonials

Community Feedback

What researchers and creators say about FireRed Image Edit

โ€œFireRed's identity consistency in v1.1 is remarkable. Face and character preservation across edits rivals closed-source solutions, and the open-source availability accelerates our research.โ€

User avatar

Dr. Wei Zhang

AI Research Scientist

Dr. Wei Zhang: โ€œFireRed's identity consistency in v1.1 is remarkable. Face and character preservation across edits rivals closed-source solutions, and the open-source availability accelerates our research.โ€

Sophia Martinez: โ€œThe multi-element fusion feature is a game-changer. Combining 10+ elements with automatic cropping and stitching saves hours of manual compositing work.โ€

Kenji Tanaka: โ€œPhoto restoration quality is outstanding. Old family photos come back to life with natural colors and sharp details. The 4.5-second inference makes batch processing practical.โ€

Emily Rogers: โ€œThe bilingual understanding is seamless. I write instructions in English, my colleague writes in Chinese, and FireRed handles both with equal precision. Truly impressive.โ€

Liu Chenxi: โ€œVirtual try-on with FireRed has transformed our product photography pipeline. Realistic garment fitting on different body types without expensive photo shoots.โ€

Anna Kowalski: โ€œThe portrait makeup capabilities cover everything from subtle beauty retouching to bold creative looks. Dozens of styles available out of the box with consistent quality.โ€

Raj Patel: โ€œTraining on 1.6 billion samples really shows. The model generalizes across diverse editing scenarios without fine-tuning. The Lightning 8-step mode is perfect for real-time applications.โ€

Yuki Nakamura: โ€œFont style reference and text rendering are best-in-class. FireRed preserves text styles with high fidelity, which is critical for our multilingual marketing materials.โ€