C

Cogvideox 5b I2V

Developed by THUDM
CogVideoX is an open-source video generation model originating from Qingying. This version is a dedicated image-to-video model based on a 5B parameter scale.
Downloads 89.01k
Release Time : 9/16/2024

Model Overview

CogVideoX-5B-I2V is a generation model specifically designed for image-to-video tasks, capable of converting static images into dynamic video sequences.

Model Features

High-Quality Video Generation
Capable of generating 6-second videos at 8 frames per second with 720x480 resolution
Multi-Precision Support
Supports various inference precisions including BF16, FP16, FP32, FP8, and INT8
Low VRAM Optimization
Can run on low VRAM GPUs through quantization techniques, requiring as little as 4.4GB VRAM

Model Capabilities

Image-to-Video Conversion
Dynamic Video Generation
Text-Prompted Video Editing

Use Cases

Content Creation
Short Video Production
Convert static images into dynamic short videos
Generate smooth 6-second videos
Education
Teaching Demonstrations
Convert teaching illustrations into dynamic presentations
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase