On September 3, Jiemeng AI officially launched on Volcano Engine, offering API services to enterprises, covering three categories: image generation, video generation, and digital human generation. The Jiemeng AI models released this time include the latest online versions that are popular among end-users, such as Text-to-Image 3.0, Text-to-Image 3.1, and Image-to-Image 3.0. These models have been finely optimized from multiple dimensions based on actual user creation feedback and in-depth understanding, making them more suitable for real-world creative needs and helping enterprises quickly build their own "creative production line".
Jiemeng AI Text-to-Image 3.0 is specifically enhanced for poster design scenarios, strengthening text and image layout, artistic fonts, and text accuracy, generating content that better meets commercial design requirements, easily producing marketing materials with a unified style and professional layout. Text-to-Image 3.1 builds upon the advantages of its predecessors, further improving overall visual aesthetics and scene richness, upgrading color, lighting, and composition comprehensively, presenting a cinematic-level visual quality. Image-to-Image 3.0 has strong editing capabilities, supporting precise image editing based on text instructions, capable of accurately executing local editing instructions such as "background replacement" and "title adjustment," while maintaining facial features and detailed textures of the person. In real image processing and poster text and image creation scenarios, this capability effectively solves the problem of traditional AI image editing causing distortion, enabling efficient implementation of creative modifications.
Jiemeng AI Video 3.0 balances quality and speed, supporting professional-level controlled camera movements and 1080P high-definition rendering, achieving smooth and natural visual presentation for both product demonstration videos and self-media content. Video 3.0 Pro introduces multi-shot storytelling capabilities, allowing seamless transitions between shots based on text instructions, with smooth movement of body language and object motion paths. It also supports two modes: "pure text generation" and "first frame image + text," adapting to diverse creative scenarios.
Jiemeng AI Digital Human focuses on usability, launching a single-image digital human approach driven by audio or video. Users only need one image, an audio clip, or a video to instantly bring characters, cartoons, or virtual figures to life. This capability precisely matches sounds, expressions, mouth shapes, and body movements, while maintaining the original image's main characteristics. Whether it's realistic people or anime characters, it can achieve vivid, expressive, and dynamic performances, showing versatility and performance, with one "person" having countless faces.
Jiemeng AI partners with Volcano Engine to provide cost-effective API services for enterprises, helping them efficiently implement creativity. New users can enjoy discounts as low as 30%, with a major promotion available exclusively in September.
