On July 14, 2026, Cheer Holding, Inc. (CHR.US), a leading mobile internet infrastructure and platform service provider, announced that its self-developed AI portrait and digital identity creation platform, Klon AI, has officially concluded its invite-only beta testing and is now fully available worldwide. Users can download the app via the App Store or Google Play.
Following internal testing in North America, Latin America, Japan, South Korea, and Southeast Asia, Klon AI has become the world's first consumer-grade application to deeply integrate AI visual generation with personal digital image management.
Key Product Features and Highlights
Users can generate multi-style portraits of professional studio quality and build a sustainable, evolving digital avatar by simply uploading 3-5 personal photos, eliminating the need for professional equipment or photography teams. The official release opens four core modules: a multi-style portrait generation engine, a global scene library with thousands of templates, a digital avatar system that maintains identity consistency across different scenarios, and an intelligent content tool that automatically transforms static portraits into social media short videos, forming a complete loop from creation to distribution.
The product's key differentiator is the cross-scenario consistency of the digital avatar. This ensures a user's facial features and personal style remain unified across various mediums like portraits, short videos, and virtual avatars, solidifying them as a recognizable personal digital asset.
Technical Leadership and Advantages
Klon AI focuses on deep engineering implementation of proven, scalable technologies rather than pursuing unproven concepts. Its technical advantages include high-quality generation using a latent diffusion model base with a DiT architecture for enhanced detail, skin texture, and composition stability. Identity consistency is achieved by fusing InstantID / IP-Adapter-FaceID for training-free identity injection with lightweight LoRA fine-tuning, validated by ArcFace facial embeddings for similarity, enabling second-level avatar creation.
Composition and image quality are precisely controlled via ControlNet, with clarity optimized using mature face restoration and super-resolution techniques. The static-to-video conversion employs a proven image-to-video diffusion scheme to create smooth short videos while preserving identity. Speed and cost are managed through LCM / Turbo distillation acceleration, model quantization, and elastic cloud inference, compressing single-image generation latency to the second level. Internal test data shows the digital avatar maintains over 95% identity feature consistency across continuous creations.
Ease of Use and Product Coverage
The complex technology stack is fully encapsulated within a simple interface, requiring just three steps: upload source material, select a style and scene, and generate with one click—no AI or photography knowledge needed. The product supports both iOS and Android with cloud-synced assets. It offers multi-language interfaces and localized templates in multiple regions, covering scenarios such as personal portraits, social content, professional profiles, e-commerce displays, and virtual avatars to meet the diverse needs of individual creators and light commercial users.
Klon AI will continue to iterate its model performance based on real usage data and plans to gradually introduce more enhanced capabilities in future versions. This approach combines mature generative AI technology with personal digital identity management, with a strong focus on engineering practicality and deployability.