Video Generation

# Video Generation in HARPA AI

The default HARPA AI CloudGPT connection comes with built-in video generation. You can generate videos from a text prompt or by providing a reference image along with a prompt — all directly in the chat.

# How to Use Video Generation

1. Make sure you have a HARPA AI CloudGPT connection selected.

2. Ask the AI to generate a video, e.g. "Generate a video of a cat playing piano" or "Create a short clip of waves crashing on rocks."

   Or provide a reference image and describe what should happen in the video, e.g. "Animate this image — make the clouds move slowly across the sky."

You can download the generated video using the player menu.

# Prompt Examples

The AI understands detailed descriptions — experiment with camera movements, lighting, and styles:

From text prompt:

Cinematic drone shot slowly rising above a foggy mountain forest at sunrise, golden light rays piercing through the trees, smooth upward camera movement, 5 seconds

Time-lapse of a flower blooming in a dark studio, single spotlight illuminating petals unfolding, shallow depth of field, macro lens perspective

From image + prompt:

Animate this scene — add gentle wind blowing through the grass, clouds drifting slowly in the background, birds flying across the sky

Make the water in this photo come alive — add flowing ripples, subtle reflections shimmering, and leaves gently floating downstream

Stylized & artistic:

Ink wash painting coming to life — a dragon slowly unfurling from a scroll, traditional Chinese brush strokes animating fluidly, muted earth tones with red accents

Retro VHS footage of a neon-lit city street at night, slight camera shake, analog scanlines and color bleeding, 80s synthwave atmosphere

Feel free to mix and match these parameters to get the result you want.