Wan 2.5 Text to Video
Wan2.5-Preview generates synchronized multimedia content, merging text, image, video, and audio seamlessly.
API
If you're looking for an API, you can choose from your desired programming language.
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
import requests
api_key = "YOUR_API_KEY"
url = "https://api.segmind.com/v1/wan-2.5-t2v"
# Prepare data and files
data = {}
files = {}
data['seed'] = 42
data['size'] = "1280*720"
# 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
data['prompt'] = "A vibrant young woman in her early 20s rushes toward the camera in Shibuya Crossing at night, ecstatic and wide-eyed, shouting passionately into a sleek silver microphone. She wears a bright orange bomber jacket and has wireless earbuds in her ears. She yells: “wan2.5 is live on Segmind — insane detail and flow, you gotta try it today!” rain-slicked crosswalks, glowing neon red-cyan-yellow billboards, blurred crowds in motion, cinematic handheld follow-shot, sharp face focus, shallow depth of field. 4K UHD, bold saturated tones, viral UGC vibe."
data['duration'] = "5"
data['negative_prompt'] = "No rainy scenes."
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 imageAttributes
Sets random seed for consistency. Use 42 for default or randomize for new results.
Determines video resolution. For high-quality use 1920x1080; for quick previews use 832x480.
Allowed values:
Upload audio for sync. Use a short loop for repetitive elements or music for dynamic output.
Input descriptive text for generation. Use vivid imagery for creative outputs or concise text for focused results.
Sets video length. Use 5s for short notes, 10s for comprehensive views.
Allowed values:
Avoid specific elements in output. Mention unwanted elements for clarity.
Toggle prompt enhancement. Use true for richer outputs or false for original input only.
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 Text to Video
# Wan2.5-Preview: Effective Usage Guide
Wan2.5-Preview is a multimodal powerhouse that unifies text, image, video and audio generation into one seamless system. Follow this guide to harness its full potential across a range of creative and professional use cases.
## Getting Started
1. **Prompt** (required): Craft a vivid, descriptive text prompt—e.g., “A bustling cyberpunk market at dusk, neon signs reflecting on wet pavement.”
2. **Seed** (optional): Use `42` for reproducible outputs or pick a random integer for fresh variations.
3. **Size** (optional): Select resolution from `832x480` up to `1920x1080`.
4. **Duration** (optional): Choose `5s` for quick loops or `10s` for richer scenes.
5. **Audio** (optional): Supply a URL to an MP3—music, dialogue or sound effects.
6. **Negative_Prompt** (optional): Specify unwanted elements, e.g., “No crowds.”
7. **Enable_Prompt_Expansion** (optional): `true` for more detail, `false` for strict adherence.
## Parameter Recommendations by Use Case
### Content Creation
- Size: `1920x1080`
- Duration: `10s`
- Audio: Upbeat track URL
- Seed: Fixed (e.g., `42`) for series consistency
- Expansion: `true`
### Digital Marketing
- Size: `1280x720`
- Duration: `5s` for quick ads
- Audio: Short, catchy loop
- Negative Prompt: “No logos” to avoid clashes
- Expansion: `true`
### Game Development
- Size: `1920x1080` for cinematics, `832x480` for thumbnails
- Duration: `10s`
- Seed: New random seeds for varied prototypes
- Prompt: Include style reference, e.g., “in Unreal Engine cinematic style”
### UI/UX Design
- Size: `1280x720` or portrait `720x1280`
- Duration: `5s` for micro-interactions
- Prompt: “Generate a sleek dashboard with pastel color scheme”
- Expansion: `false` for precise layouts
### Film Production
- Size: `1920x1080`
- Duration: `10s`
- Audio: Upload voice-over and ambient bed
- Negative Prompt: “No watermarks”
- Seed: Keep consistent across shots
### Educational Content
- Size: `1280x720`
- Duration: `10s`
- Prompt: “Animate the water cycle with labeled stages”
- Audio: Narration URL
- Expansion: `true`
## Prompt Crafting Tips
- Use concrete nouns and vivid adjectives.
- Reference artists or styles for consistent aesthetics.
- Leverage negative prompts to exclude unwanted details.
- Enable prompt expansion to enrich context and depth.
By adjusting these parameters, Wan2.5-Preview will deliver high-fidelity 1080p videos, synchronized audio, and stunning images tailored to your project’s needs. Happy creating!```
Other Popular Models
Discover other models you might be interested in.
sdxl-img2img
SDXL Img2Img is used for text-guided image-to-image translation. This model uses the weights from Stable Diffusion to generate new images from an input image using StableDiffusionImg2ImgPipeline from diffusers
sdxl-inpaint
This model is capable of generating photo-realistic images given any text input, with the extra capability of inpainting the pictures by using a mask
sd1.5-majicmix
The most versatile photorealistic model that blends various models to achieve the amazing realistic images.
sd2.1-faceswapper
Take a picture/gif and replace the face in it with a face of your choice. You only need one image of the desired face. No dataset, no training
