How to Use Runway ML: AI Video Generation and Editing Guide

Runway ML is the leading AI video studio — used by professional filmmakers, YouTubers, and marketers to create and edit video with AI. This guide covers the main features and how to get good results.

Runway Plans and Credits

Free: 125 credits, watermarked videos, limited resolution
Standard ($12/month): 625 credits, no watermark, 1080p
Pro ($28/month): 2250 credits, max resolution, 4K upscaling
Unlimited ($76/month): Unlimited generations, full feature access

Credit costs: Gen-3 video costs 5 credits per second of output.

Gen-3 Alpha: Text-to-Video

Runway’s most capable generation model.

Accessing Gen-3

Log into runwayml.com
Click Generate → Text/Image to Video
Select model: Gen-3 Alpha (highest quality) or Gen-3 Alpha Turbo (faster, cheaper)

Prompting for Gen-3

Basic structure:

[Camera action], [Subject], [Action/Movement], [Setting], [Style/Mood]

Examples:

Low angle tracking shot, a golden retriever running through a field of sunflowers, 
slow motion, warm afternoon light, cinematic

Extreme close-up, water droplets falling on a rose petal, macro lens, 
soft bokeh background, nature documentary style

Aerial drone shot slowly pushing forward, a coastal town at sunset, 
small boats in the harbor, warm golden light, photorealistic

Camera Movements

Specify camera movement for professional-looking results:

Static shot — camera doesn’t move
Slow zoom in / slow zoom out
Pan left / pan right
Tracking shot — follows the subject
Dolly forward / dolly backward
Crane shot — camera rises or falls
Handheld — natural, slightly shaky movement
Aerial drone shot — bird’s eye view

Style Keywords

Cinematic: anamorphic lens, film grain, shallow depth of field, cinematic color grading

Documentary: handheld camera, natural lighting, observational style

Commercial: clean and bright, professional lighting, advertising aesthetic

Artistic: impressionistic, painterly, stylized, abstract

Image-to-Video

Animate a still image — one of Runway’s most useful features.

Click Image to Video
Upload your source image
Add a motion prompt describing what should happen

Motion prompt tips:

The figure slowly turns their head to look at the camera

Clouds drift across the sky, light shifts gradually

Leaves flutter gently in a breeze, subtle movement throughout

What works well:

Natural, subtle motions
Environmental movement (water, wind, light changes)
Simple actions (person walking, turning, gesturing)

What doesn’t work well:

Extreme motion (too much distortion)
Text (distorts heavily)
Multiple disconnected subjects moving independently

Video Editing Features

Background Removal (Remove BG)

Upload your video
Select Remove Background
Runway isolates the subject automatically
Replace background with any image, video, or solid color

Best for: Green screen alternative, social media content, product videos

Inpainting (Erase and Replace)

Remove unwanted objects from video:

Upload video
Select Inpaint
Paint over the object you want to remove
Runway fills it with generated content

Best for: Removing microphones, logos, people, objects from footage

Motion Blur

Add professional motion blur to text or graphics moving across the frame.

Infinite Image (Outpainting)

Extend the edges of an image beyond its original boundaries — useful for creating wider or taller compositions from existing footage.

Professional Workflows

Generate concept video (10 seconds) with Gen-3
Remove background if needed
Export at 1080p
Add music in your editing software
Add text overlay (Runway has basic text tools, or use Canva/Premiere)

Product Showcase Video

Upload product photo
Use Image-to-Video with subtle rotation/orbit motion
Use background removal to isolate product
Place on custom background (gradient, studio, lifestyle setting)
Add zoom and text in post

Film and Creative Projects

Directors use Runway for:

Previsualization: Generate rough versions of scenes before expensive production
Concept art animated: Turn still concept art into motion
VFX shots: Generate impossible shots as a base for compositing
Storyboarding: Generate visual storyboard frames quickly

Getting Consistent Characters

One challenge with video generation: characters change between clips.

Workaround:

Generate your subject in a reference image first (Midjourney, Stable Diffusion)
Use that same image as the source for image-to-video
Keep the same image as reference across multiple clips

For series content requiring the same character, this isn’t fully solved yet — character consistency across clips requires significant post-production or using purpose-built character consistency tools.

Comparing Gen-3 vs. Gen-3 Turbo

Feature	Gen-3 Alpha	Gen-3 Alpha Turbo
Quality	Higher	Good
Speed	~90 seconds	~30 seconds
Cost	5 credits/second	2.5 credits/second
Best for	Final output	Iteration/testing

Use Turbo for exploring ideas, Gen-3 for final deliverables.

Runway vs. Alternatives

Runway Gen-3: Best for professional filmmakers and commercial production. Most editing tools.

Kling AI: Better motion quality and longer clips at lower cost. Less editing tooling.

Pika: Easier to use, good for quick social content, lower quality ceiling.

Sora (OpenAI): Excellent quality but limited availability and less editing control.

Recommendation: Start with Kling for cost-effective generation, use Runway when you need the editing suite.

Tips for Better Results

Be specific about camera work: “Cinematic close-up tracking shot” beats “show a person”

Describe lighting: “Soft afternoon backlight” or “harsh neon lighting” dramatically changes the mood

Test with Turbo first: Use the faster/cheaper model to iterate on prompts, then generate final with Gen-3

Short clips look better: 4-8 second clips have better motion consistency than 10 seconds

Combine with traditional editing: Runway generates, a tool like Premiere Pro or DaVinci Resolve finishes the job