Wan 2.5 Image to Video

Wan2.5-Preview creates stunning, high-resolution videos with flawless audio synchronization from multiple inputs.


API

If you're looking for an API, you can choose from your desired programming language.

POST
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 import requests api_key = "YOUR_API_KEY" url = "https://api.segmind.com/v1/wan-2.5-i2v" # Prepare data and files data = {} files = {} data['seed'] = 42 # For parameter "audio", you can send a raw file or a URI: # files['audio'] = open('IMAGE_PATH', 'rb') # To send a file data['audio'] = 'null' # To send a URI # For parameter "image", you can send a raw file or a URI: # files['image'] = open('IMAGE_PATH', 'rb') # To send a file data['image'] = 'https://segmind-resources.s3.amazonaws.com/output/21aeb463-bb17-4536-864b-0bd1e11594a9-EMXN1y8qTgoGdXBsb2FkEg55bGFiLXN0dW50LXNncBo0YWlfcG9ydGFsLzE3NTM5NjM5NTIvbjdwNDlYOURCbS8yZTk0X2wwXzAwMS0wXzAuanBlZw_1000x1000.webp' # To send a URI data['prompt'] = "Kitten in a McDonald's uniform stands on a stool, grilling burger patties. It flips the patties with a spatula, watches them sizzle, and occasionally looks around while steam rises from the grill." data['duration'] = "5" data['resolution'] = "720p" data['negative_prompt'] = "unnecessary clutter, dark shadows" data['enable_prompt_expansion'] = True headers = {'x-api-key': api_key} # If no files, send as JSON if files: response = requests.post(url, data=data, files=files, headers=headers) else: response = requests.post(url, json=data, headers=headers) print(response.content) # The response is the generated image
RESPONSE
video/mp4
HTTP Response Codes
200 - OKImage Generated
401 - UnauthorizedUser authentication failed
404 - Not FoundThe requested URL does not exist
405 - Method Not AllowedThe requested HTTP method is not allowed
406 - Not AcceptableNot enough credits
500 - Server ErrorServer had some issue with processing

Attributes


seedint ( default: 42 )

Sets a random seed for consistent outputs. Use values between 1 and 100 for variation.


audiostr ( default: 1 )

Upload an audio file for syncing. Use a song clip or melody for dynamic results.


imagestr *

Input image for video generation. Choose a high-resolution image for best quality.


promptstr *

Text description for video creation. Include vivid visuals for creative animations.


durationenum:str ( default: 1 )

Sets the video length. Choose 5 seconds for shorter clips and 10 for longer scenes.

Allowed values:


resolutionenum:str ( default: 720p )

Set video quality. Use 1080p for high-quality renders and 480p for faster results.

Allowed values:


negative_promptstr ( default: unnecessary clutter, dark shadows )

Avoid certain elements in generation. Include unwanted objects or colors for exclusion.


enable_prompt_expansionbool ( default: true )

Activates prompt optimizer for enhanced results. Set to true for more detailed outputs.

To keep track of your credit usage, you can inspect the response headers of each API call. The x-remaining-credits property will indicate the number of remaining credits in your account. Ensure you monitor this value to avoid any disruptions in your API usage.

Resources to get you started

Everything you need to know to get the most out of Wan 2.5 Image to Video

# Wan2.5-Preview: Guide to Effective Use

Wan2.5-Preview is a state-of-the-art multimodal AI model that generates up to 10-second 1080p videos with synchronized audio, image input, and rich textual prompts. Follow this guide to optimize parameters and achieve cinematic-quality outputs across diverse applications.

## 1. Getting Started  
1. Provide a high-resolution `image` URL (recommended 1080×1080 px or above).  
2. Write a concise, vivid `prompt` describing scene composition, motion, and style.  
3. (Optional) Upload an `audio` file (MP3/ WAV) to drive soundtrack or narration.  
4. Adjust advanced parameters for duration, resolution, seed, and negative prompts.

## 2. Recommended Parameter Presets  

Use Case                | Resolution | Duration | Seed | Prompt Expansion | Negative Prompt  
----------------------- | ---------- | -------- | ---- | ---------------- | ----------------  
Professional Ads        | 1080p      | 10s      | 12   | true             | “grain, low-contrast”  
Social Media Teasers    | 720p       | 5s       | 42   | true             | “blurry edges, dark shadows”  
Educational Narration   | 720p       | 10s      | 7    | true             | “background noise, clutter”  
Music Video Snippets     | 1080p      | 10s      | 29   | false            | “static cameras, dull colors”  
Quick Prototyping       | 480p       | 5s       | 99   | false            | “extra objects, lens flare”  

- **Resolution**:  
  • 1080p for broadcast-grade quality  
  • 720p for balanced speed and detail  
  • 480p for fast previews  
- **Duration**: 5 s for short loops; 10 s for extended scenes.  
- **Seed (1–100)**: Ensures reproducibility. Reuse same seed + inputs for consistent results.  
- **Enable Prompt Expansion**: Improves detail but may increase processing time. Set to `true` for complex scenes.

## 3. Prompt Engineering Tips  
- Start with the subject: “A sleek drone sweeping over neon-lit cityscapes…”  
- Specify camera movements: “slow dolly, zoom out, pan left to right.”  
- Include mood, lighting, color palette: “warm sunset glow, soft shadows.”  
- Use `negative_prompt` to exclude unwanted elements: e.g., “no rain, no subtitles.”  
- Combine keywords and short phrases—avoid run-on sentences.

## 4. Audio Synchronization  
- Upload voice-over or music via the `audio` field.  
- Ensure audio length matches video duration to prevent cuts.  
- For multi-track sound design, overlay effects (e.g., “wind whoosh,” “crowd murmur”) in a single file.

## 5. Best Practices  
- Test with lower resolution (480p) when crafting prompts.  
- Lock your seed before final render.  
- Review generated thumbnails before full download.  
- Tweak negative prompts iteratively to refine outputs.

By following these presets and tips, you’ll harness Wan2.5-Preview’s full potential for polished, synchronized audiovisual content every time.

Other Popular Models

Discover other models you might be interested in.

Cookie settings

We use cookies to enhance your browsing experience, analyze site traffic, and personalize content. By clicking "Accept all", you consent to our use of cookies.