AI-Generated Video

Create a video entirely from AI-generated content: images, animation, and narration.

Goal

Generate images from text prompts, animate them into video segments, add AI-generated speech, and combine everything with transitions.

Full Example

import asyncio
from videopython.base import Video, FadeTransition, Resize, TransformationPipeline
from videopython.ai import TextToImage, ImageToVideo, TextToSpeech

async def create_ai_video():
    # Define scenes with prompts
    scenes = [
        {
            "image_prompt": "A serene mountain landscape at sunrise, photorealistic",
            "narration": "In the mountains, every sunrise brings new possibilities.",
        },
        {
            "image_prompt": "A flowing river through a forest, cinematic lighting",
            "narration": "Nature flows with endless energy and grace.",
        },
        {
            "image_prompt": "A starry night sky over a calm lake, dramatic",
            "narration": "And when night falls, the universe reveals its wonders.",
        },
    ]

    # Initialize AI generators
    image_gen = TextToImage(backend="openai")
    video_gen = ImageToVideo(backend="local")
    speech_gen = TextToSpeech(backend="openai")

    videos = []
    for scene in scenes:
        # Generate image
        image = await image_gen.generate_image(scene["image_prompt"])

        # Animate image to video (4 seconds)
        video = await video_gen.generate_video(image=image, fps=24)

        # Resize to consistent dimensions
        video = Resize(width=1920, height=1080).apply(video)

        # Generate narration audio
        audio = await speech_gen.generate_audio(scene["narration"])

        # Add audio to video segment
        video.add_audio(audio)
        videos.append(video)

    # Combine all segments with fade transitions
    fade = FadeTransition(effect_time_seconds=1.0)
    final = videos[0]
    for next_video in videos[1:]:
        final = fade.apply((final, next_video))

    return final

# Run and save
video = asyncio.run(create_ai_video())
video.save("ai_generated.mp4")

Step-by-Step Breakdown

1. Generate Images

image_gen = TextToImage(backend="openai")  # Uses DALL-E 3
image = await image_gen.generate_image("A serene mountain landscape at sunrise")

Available backends:

Backend	Model
openai	DALL-E 3
local	Stable Diffusion XL

2. Animate to Video

video_gen = ImageToVideo(backend="local")  # Uses Stable Video Diffusion
video = await video_gen.generate_video(image=image, fps=24)

Local Models

Local backends (ImageToVideo, TextToVideo) require significant GPU memory. Expect 8GB+ VRAM for smooth operation.

3. Generate Speech

speech_gen = TextToSpeech(backend="openai")  # Uses OpenAI TTS
audio = await speech_gen.generate_audio("Your narration text here")

Available backends:

Backend	Model
openai	OpenAI TTS
elevenlabs	Multilingual v2
local	Bark

4. Combine Segments

fade = FadeTransition(effect_time_seconds=1.0)
final = fade.apply((video1, video2))

Tips

• Consistency: Use similar prompt styles across scenes for visual coherence.
• Timing: Match narration length to video segment duration.
• Quality: OpenAI backends generally produce higher quality but cost money. Local backends are free but require GPU.