Runway Core Workflow B

Overview

Image-to-video and video-to-video generation: animate still images and transform existing videos.

Prerequisites

• Completed runway-core-workflow-a

Instructions

Step 1: Image-to-Video

from runwayml import RunwayML
client = RunwayML()

# Animate a still image
task = client.image_to_video.create(
    model='gen3a_turbo',
    prompt_image='https://your-cdn.com/landscape.jpg',  # URL to source image
    prompt_text='Camera slowly pans right revealing mountains, gentle wind in trees',
    duration=5,
)
result = task.wait_for_task_output()
print(f"Animated video: {result.output[0]}")

Step 2: Image-to-Video with Data URI

import base64

# Load local image as data URI
with open('photo.jpg', 'rb') as f:
    image_data = base64.b64encode(f.read()).decode()
    data_uri = f"data:image/jpeg;base64,{image_data}"

task = client.image_to_video.create(
    model='gen3a_turbo',
    prompt_image=data_uri,
    prompt_text='Subtle motion, gentle camera push in, atmospheric lighting',
    duration=5,
)

Step 3: Video-to-Video (Style Transfer)

# Transform an existing video with a new style
task = client.video_to_video.create(
    model='gen3a_turbo',
    prompt_video='https://your-cdn.com/input-video.mp4',
    prompt_text='Transform to watercolor painting style, soft colors, artistic brushstrokes',
)
result = task.wait_for_task_output()
print(f"Styled video: {result.output[0]}")

Step 4: Image Specifications

Supported formats: JPEG, PNG, WebP
Supported resolutions:
  - Gen-3 Alpha Turbo: 1280x768 or 768x1280
  - Input images are automatically resized
Max file size: 16MB (URL), varies for data URI

Output

• Still images animated with motion prompts
• Local images encoded as data URIs
• Videos restyled with text prompts
• Proper image format handling

Error Handling

Resources

• Runway API Reference
• Input Parameters
• SDKs

Next Steps

Error handling: runway-common-errors