Dear SeaArt.AI User,

This update introduces multiple video models and adds new image editing features. Details are as follows:

✨ New Models Launched ✨

1. Video Model: SeaArt Sono Epic

▪ Model Features:

▪ Breaking Duration Limits: The first model to support ultra-long videos of up to 20 seconds, with flexible 15s and 20s options.

▪ Ultimate Character Consistency: Exceptional consistency across actions and scenes, achieving highly unified character facial features and clothing details.

▪ 🔗Model Link: https://www.seaart.ai/models/detail/d5sbt9te878c73dtueq0

2. Video Model: SeaArt Sono Wave

▪ Model Features:

▪ Multi-Shot Logical Expression: Excellent multi-angle transition handling capabilities with smooth and natural shot connections, ensuring logical consistency in storytelling.

▪ Flexible Duration, Outstanding Efficiency: Supports 5, 10, and 15-second duration customization, balancing creative freedom with exceptional cost-effectiveness.

▪ 🔗Model Link: https://www.seaart.ai/models/detail/d5u6hf5e878c73ebul9g

3. Video Model: SeaArt Opera

▪ Model Features:

▪ Cinematic Multilingual Visual Output: Supports native 1080p HD quality, compatible with Chinese, Japanese, and English output, precisely capturing and reproducing diverse global artistic styles to produce studio-grade professional visual masterpieces.

▪ Immersive Artistic Reconstruction: Achieves a sensory breakthrough of "vivid sound and color" with audio-visual synchronous generation technology, allowing music melodies and visual imagery to perfectly interweave at the moment of generation, producing highly stylistically impactful audiovisual works without post-production.

▪ 🔗Model Link: https://www.seaart.ai/models/detail/d606fl5e878c73ejm1kg

4. Video Model: Vidu Q3 pro

▪ Model Features:

▪ Professional Long Video Narrative: Supports 15-second continuous video generation, providing ample narrative space for rhythmic storylines and nuanced long-take expressions.

▪ Full-Process Shot and Text Coordination: Integrates industrial-grade multi-shot switching and native subtitle rendering technology, enabling precise control of complex cinematic language through prompts, from camera direction to subtitle embedding in one seamless flow, revolutionizing creative efficiency.

▪ 🔗Model Link: https://www.seaart.ai/models/detail/d606ft5e878c73ejnh3g

🌊 Feature Updates and Optimizations 🎨

1. LoRA Training Optimizations

▪ Model Training -> Image LoRA Training -> Z Image now includes the base model "Z Image Base."

▪ Z Image model's advanced settings now include network settings, supporting adjustment of Network Size and Network Alpha.

2. The character limit for community comments has been increased to 1,000, allowing for more in-depth exchanges of creative insights.

📱 Updates for Mobile H5 Only 📱

1. New Image Editing Feature

▪ How to Access: Workflow -> the Camera icon

▪ Supports quick editing after image upload, including popular features such as partial repainting, upscaling, and smart background removal.

2. New AI Face Swap Feature

▪ How to Access: AI Apps -> AI Face Swap

▪ Supports face swapping in images, videos, and GIFs, allowing you to quickly upload images and complete your creations.

🔧 Bug Fixes 🔧

1. Fixed an issue where black images abnormally appeared when generating images.

2. Fixed an issue where other users' avatars abnormally displayed as default avatars.

3. Fixed an issue where generated works contained noise when two self-trained LoRAs were used simultaneously.

4. Fixed an issue where some content in system notifications displayed as blank.

5. Fixed an issue where some video posts abnormally displayed as image posts on the mobile H5 version's profile page.

We invite you to try out these updates. If you have any suggestions or feedback, please leave a comment below. We will continue to gather your input and work towards providing a better, more optimized experience. →

SeaArt.AI Team