Zhipu Qingying and AI Character Creation Platform 'NianTa' Initiate Multimodal Collaboration to Promote IP Consistency Implementation

AIbase基地

Published inAI News · 4 min read · Feb 20, 2025

129

On February 19, Beijing Zhipu Huazhang Technology Co., Ltd. announced a strategic partnership between its multimodal creation platform "Qingying" and the AI character creation platform "Nieta". Together, they aim to explore the complete path from AI character design to animation creation and promote the technical implementation of IP consistency. This collaboration is designed to deeply explore the needs of the two-dimensional user base and AI application scenarios, providing users with higher quality generation results and faster inference speeds.

"Qingying" is based on Zhipu's latest upgraded video generation model, CogVideoX-2, which has undergone comprehensive updates in model structure, training methods, and data engineering, significantly enhancing the basic model capabilities for image-to-video by 38%. This model supports substantial movement of the main subject in the scene while maintaining stability, and it excels in instruction compliance and artistic style management. In this collaboration, "Qingying" launched a feature on the "Nieta" platform that generates high-quality videos based on text and images, achieving over 40,000 video generations and over 100 million tokens used on the first day of launch.

WeChat Screenshot_20250220085319.png

"Nieta" is a leading AI character creation platform with over a million monthly active users, where users can participate in scripts, experience murder mystery games, create novels and comics, and even engage in social interactions. This collaboration allows "Nieta" users to generate high-quality video content with simple text or image prompts, greatly lowering the barrier to content creation and truly achieving "one prompt, one video".

This joint initiative represents the successful implementation of Zhipu's multimodal models in commercial scenarios, marking rapid growth for Zhipu in the field of multimodal commercialization. Since the end of 2024, the usage of Zhipu's multimodal series models has increased by over 1000% on average daily compared to 2024, providing strong technical support to more industry partners and promoting the deep application of AI in content creation.

Free Trial! Zhipu Launches AI Video Product Qingying 2.0 Fully Online at Zhipu Qingyan

Qingying 2.0 has undergone a comprehensive upgrade in model architecture, training methods, and data engineering, with a significant improvement of 38% in the base model's capabilities. This enhancement enables Qingying 2.0 to generate video content with more natural movements, exquisite visuals, and stable control. Users only need to input simple prompts to allow the main subject in the video to perform significant movements while maintaining stability. For instance, Qingying 2.0 can generate a coherent action of a man stepping out of a poster, adjusting his tie, and executing close-up shots, demonstrating its ability in motion continuity and visual stability.

Zhipu AI Announces Open Source of 'Qingying' Homogeneous Video Generation Model - CogVideoX

Zhipu AI announces the open source of its video generation model CogVideoX, aiming to promote the rapid development and application of video generation technology. The CogVideoX model is based on advanced large model technology and is capable of meeting commercial-grade application requirements. The currently open sourced CogVideoX-2B version requires only 18GB of GPU memory for inference at FP-16 precision and 40GB for fine-tuning, allowing inference with a single 4090 GPU and fine-tuning with a single A6000 GPU.

Zhizhu AI Launches Next-Generation Video Generation Model CogVideoX, Available for Free Experience via 'Qingying'

Zhipu AI has launched CogVideoX, a next-generation video generation model that enhances temporal frame relationship capturing through a 3D VAE structure and 3D RoPE module. This model integrates text, time, and space into a transformer architecture, allowing for efficient video generation that meets user needs. It is available on the Zhipu Qingyan platform for free video generation services, featuring fast generation, effective instruction adhere....

Zhipu AI Launches AI Video Generation Product 'Qingying', Now Available on Zhipu Qingyan PC/App

During its OpenDay event, Zhipu AI introduced an AI large model named Qingying, capable of generating videos from any text. Qingying can produce high-definition videos at a resolution of 1440x960 within 30 seconds. Users need only to input a text (i.e., Prompt) and choose a desired style, such as cartoon 3D, black and white, oil painting, cinematic, etc., to generate the video. Qingying is now available on the Qingyan App, accessible to all users.