Home
Xiaohongshu Open-Sources FireRed-Image-Edit v1.1 AI Model, Tackles Consistency and Complex Edits
On March 9, 2026, Xiaohongshu's Super Intelligence team officially launched the FireRed-Image-Edit v1.1 model. Arriving less than a month after v1.0, this release signals a major acceleration in the platform's iteration speed for multimodal large models.
The new version delivers deep optimization for complex tasks like ID-consistent editing, multi-element fusion, portrait enhancement, and font style reference, while preserving its predecessor's strengths. It exhibits significantly improved semantic comprehension and visual generation. From a technical standpoint, v1.1 now supports full-process optimization for training and deployment, cutting inference time to 4.5 seconds and capping memory usage at 30GB—a major boost for practical, industrial-scale implementation.

As the core engine powering Xiaohongshu's general intelligence ambitions, the Super Intelligence team has fully open-sourced the project's code, technical papers, model weights, and its complete training and inference framework. This initiative not only bolsters Xiaohongshu's internal workflows for content creation, distribution, search, and advertising but also addresses an industry gap by providing a sophisticated, open-source toolkit for fine-grained image editing.
With the global large model race now focused on deep application, the rapid advancement and open-sourcing of the FireRed series highlights how leading platforms are working to democratize multimodal technology. By shifting focus from pure model development to high-performance engineering, Xiaohongshu is forging a distinct, content-centric AI advantage, poised to accelerate the adoption of multimodal intelligence in content-driven e-commerce and social platforms.
Related article
DeepMind CEO Hassabis: I sleep six hours a day, usually feel energetic around 1 a.m.
Fortune recently featured an interview with Demis Hassabis, CEO of Google DeepMind, revealing his unconventional approach to rest and productivity. Hassabis disclosed that he sleeps very little, structuring his waking hours into two distinct work blo
OpenAI, Anthropic Vie for Market Share Despite Revenue Shortfalls
Despite recent reports suggesting OpenAI missed revenue targets, creating pressure on tech stocks this Tuesday, private AI lab investors remain resilient. Seasoned backers have confirmed they will not reduce investment despite negative media coverage
California AV Compliance: A New Era of Tickets, Geofences, and 1M Miles
Guident operates an AuveTech shuttle in South Florida, managing a four-mile route in West Palm Beach and a one-mile route in Boca Raton using its remote monitoring technology. | Credit: GuidentCalifornia is redefining the regulatory landscape for dri
Related Special Topic Recommendations
Comments (0)
0/500
On March 9, 2026, Xiaohongshu's Super Intelligence team officially launched the FireRed-Image-Edit v1.1 model. Arriving less than a month after v1.0, this release signals a major acceleration in the platform's iteration speed for multimodal large models.
The new version delivers deep optimization for complex tasks like ID-consistent editing, multi-element fusion, portrait enhancement, and font style reference, while preserving its predecessor's strengths. It exhibits significantly improved semantic comprehension and visual generation. From a technical standpoint, v1.1 now supports full-process optimization for training and deployment, cutting inference time to 4.5 seconds and capping memory usage at 30GB—a major boost for practical, industrial-scale implementation.

As the core engine powering Xiaohongshu's general intelligence ambitions, the Super Intelligence team has fully open-sourced the project's code, technical papers, model weights, and its complete training and inference framework. This initiative not only bolsters Xiaohongshu's internal workflows for content creation, distribution, search, and advertising but also addresses an industry gap by providing a sophisticated, open-source toolkit for fine-grained image editing.
With the global large model race now focused on deep application, the rapid advancement and open-sourcing of the FireRed series highlights how leading platforms are working to democratize multimodal technology. By shifting focus from pure model development to high-performance engineering, Xiaohongshu is forging a distinct, content-centric AI advantage, poised to accelerate the adoption of multimodal intelligence in content-driven e-commerce and social platforms.
DeepMind CEO Hassabis: I sleep six hours a day, usually feel energetic around 1 a.m.
Fortune recently featured an interview with Demis Hassabis, CEO of Google DeepMind, revealing his unconventional approach to rest and productivity. Hassabis disclosed that he sleeps very little, structuring his waking hours into two distinct work blo
OpenAI, Anthropic Vie for Market Share Despite Revenue Shortfalls
Despite recent reports suggesting OpenAI missed revenue targets, creating pressure on tech stocks this Tuesday, private AI lab investors remain resilient. Seasoned backers have confirmed they will not reduce investment despite negative media coverage











