Select Language
Select Language

Wan 2.5 AI Video Generator

Create stunning AI videos with Alibaba's Wan 2.5 - Advanced text-to-video and image-to-video generation

Wan 2.5 delivers 24fps cinema-quality output with powerful dynamics, structural stability, and enhanced cinematic control powered by RLHF optimization.

Create Your Video

0/2000

Click to upload or drag and drop

JPG, PNG (Max 10MB)
Preview
0/2000

Cost: 60 Gems(Coins)

Preview

Your video will appear here

Generating your video...

Wan 2.5 Performance Benchmark

Compared to Wan 2.2, Wan 2.5's performance improvements demonstrate powerful enhancements in generation speed, video quality, and semantic compliance, while maintaining the revolutionary open-source accessibility of Wan 2.2.

🏆
+30%
Quality Improvement
+25%
Speed Enhancement
+40%
Accuracy Boost

Performance Comparison

Performance Metric Wan 2.5 Wan2.2 Improvement
Generation Speed Enhanced Baseline +25% Faster
Video Quality Improved Standard +30% Better
Semantic Compliance Advanced Good +40% Accuracy
Motion Reconstruction Superior Standard +35% Smoother
Hardware Compatibility Optimized Compatible +20% More Efficient
Open Source Access Apache 2.0 Apache 2.0 Maintained

Technical Improvements

Enhanced MoE Architecture

Optimized parameter distribution for better efficiency

🎨

Improved VAE Integration

Better compression and quality retention

Multi-GPU Optimization

Enhanced scalability and resource utilization

🚀

Apache 2.0

Maintaining open-source accessibility

What is Wan 2.5?

Wan 2.5 (Tongyi Wanxiang 2.5) is Alibaba's latest advanced AI video generation model that offers flexible resolution options (720p and 1080p) and customizable durations. It provides high-quality video generation with precise control, making it perfect for various content creation needs from social media to professional video production.

🎬 Flexible Duration Control
🎯 HD Quality Options
🎵 Synchronized A/V
💎 Fast Generation

Wan 2.5 Core Features

Native Multimodal Video Generation Platform - Complete Guide

🧠

Native Multimodal Architecture

Wan 2.5 adopts a unified understanding and generation framework that flexibly supports text, image, video, and audio input and output, achieving deep alignment through joint multimodal training.

🎵

Synchronized A/V Generation

Wan 2.5 natively supports high-fidelity, high-consistency video generation with synchronized audio, including multi-person voices, sound effects, and background music, creating immersive audio-visual experiences.

🎬

Professional Video Quality

Wan 2.5 generates 24fps cinema-quality 1080p HD videos lasting 10 seconds, with powerful dynamics, structural stability, and upgraded cinematic control system.

🎨

Advanced Image Editing

Wan 2.5 supports conversation-based image editing with pixel-level precision for multi-concept fusion, material transformation, product color change, and creative typography tasks.

RLHF Optimization

Wan 2.5 implements Reinforcement Learning from Human Feedback (RLHF), continuously aligning with human preferences to enhance image quality and video dynamics for better user satisfaction.

🔊

Rich Audio Capabilities

Wan 2.5 supports high-fidelity sound, ASMR, ambient audio, music, multi-language support, and audio-driven video generation with seamless audio-video synchronization.

Why Wan 2.5?

Flexible Duration Options

Create videos in 5s or 10s duration for Image-to-Video mode. Choose the perfect length for your content needs.

Smart Resolution Pricing

Choose 720p (60 gems) for quick previews or 1080p (120 gems) for professional quality. Intelligent pricing based on your needs and budget.

Dual Generation Modes

Switch between Text-to-Video and Image-to-Video modes. Transform your ideas or images into stunning videos with AI.

Create ads and social media clips in minutes

——— Marketers

Turn scripts into engaging videos instantly.

——— Content Creators

Generate training videos or presentations effortlessly.

——— Educators

How to use Wan 2.5?

From Words to Videos: Your Story, Powered by Wan 2.5 AI

1

Step 1

Choose between Text-to-Video or Image-to-Video mode, then enter your prompt or upload an image.

2

Step 2

Select your desired resolution (720p or 1080p) and duration (5s or 10s for Image-to-Video).

3

Step 3

Click Create Button – rendering takes 1-5 minutes.

Wan 2.5 Frequently Asked Questions

Find answers to common questions about Wan 2.5 AI video generation.

01

What is Wan 2.5?

Wan 2.5 (Tongyi Wanxiang 2.5) is Alibaba's latest AI video generation model with flexible resolution (720p/1080p) and duration options (5s/10s).

02

What are the pricing options for Wan 2.5?

Text-to-Video: 720p costs 60 gems, 1080p costs 120 gems. Image-to-Video: 720p 5s costs 60 gems, 720p 10s costs 120 gems, 1080p 5s costs 120 gems, 1080p 10s costs 200 gems.

03

What's the difference between 720p and 1080p?

720p costs 60 gems and is perfect for quick previews, while 1080p costs 120 gems and provides professional HD quality output.

04

Can I use Wan 2.5 for commercial projects?

Yes, videos created with Wan 2.5 can be used for commercial purposes.

05

How does RLHF improve Wan 2.5's performance?

Wan 2.5 implements Reinforcement Learning from Human Feedback (RLHF), continuously aligning with human preferences to enhance image quality and video dynamics for better user satisfaction. This optimization ensures Wan 2.5 delivers results that match user expectations and creative vision.

06

What types of audio can Wan 2.5 generate?

Wan 2.5 supports high-fidelity sound generation including ASMR, ambient audio, music, and offers multi-language support. Wan 2.5 also features audio-driven video generation with seamless audio-video synchronization, making it a comprehensive solution for multimedia content creation.