Two people wearing VR headsets interact with a large, illuminated digital screen in a modern, colorful room.

AI video editing using Runway: Part Two

From text to detailed image generation prompt.

Posted by By pixel perfect 3 Min Read June 6, 2025

We continue our series introducing one of the world’s most popular online generative AI video production toolkits, as we create a short video for a brand new immersive experience.

The AI tools and models used in this series

We’ll use runway.ml to generate video clips and assemble them into a sequence, and we’ll use additional AI tools to help with initial ideation and to create text prompts to use at runway to help generate high quality video clips.

The main steps in the project:

  1. Generate an overview for an ‘immersive experience’ promotional video using perplexity.ai
  2. Generate still images using the flux-1.1-pro model at replicate.com to use as ‘first frame’ prompts at runway
  3. Generate short video clips using runway’s latest Gen-4 model
  4. Use runway’s online video editor to sequence the promotional video
  5. Use the speech-02-hd AI model at replicate.com to generate a realistic-sounding voiceover
  6. Use udio.com to generate a background music track
  7. Add text titles and export the video file

Part 2: Convert the suggested shots into text prompts to generate images.

The latest runway video generation models can use the combination of a reference image AND a text prompt to create short video clips.

In perplexity, we asked for the suggested opener ‘hook’ shot to be converted into a detailed text prompt ready to use to generate a still image:

A computer screen displays a chat interface with a prompt and an AI-generated descriptive text about an immersive environment.

In runway, this newly-created text prompt will be used to generate four ‘candidate images’. The runway dashboard provides a clear place to start:

A dashboard interface for Runway displays options to generate videos, animate characters, and generate images.

Since using runway last year, the video generation models have been upgraded to include ‘Gen-4’ and ‘Gen-4 turbo’:

Dropdown menu showing options for Gen-4 Turbo, Gen-4, Gen-3 Alpha Turbo, and Gen-3 Alpha AI models for image or video generation.

These newer models can use a prompt image AND prompt text to generate video clips:

Screenshot of an image upload interface prompting users to drop an image or create one, with asset selection options below.

Click on ‘create image’, then paste in the image prompt. Click ‘Generate’ to create four ‘candidate images’:

A computer screen displays AI-generated images and a text prompt describing a dynamic, tech-themed digital scene.

It’s worth reviewing the image generation settings. We raised the ‘aesthetic range’ slider to the maximum, hence the ‘diverse’ images it’s created:

A dialog box showing "Aesthetic Range" set to 5, with a description about increasing creative variation.

A closer look at the four generated images:

A woman in a futuristic outfit walks toward a bright portal surrounded by colorful, crystalline structures.
A person uses augmented reality glasses, interacting with digital projections in a dark, modern exhibition space.
A person wearing a VR headset views a colorful, futuristic space scene with spaceships and swirling clouds.
A woman in futuristic clothing interacts with a large transparent digital screen in a high-tech environment.

Generating the four images has consumed 32 credits, which equates to around £0.50 or 60 cents at time of writing. This could add up very quickly if you don’t like the initial results.

Here’s runway’s ‘credits tariff’ in June 2025:

A table displays the credit cost per second or generation for various Gen-4 AI video and image tools.

In the next part of this series…

We’ll use a dedicated image generation model to generate additional still images.