How to Use Runway Gen-5 to Generate Video From a Prompt
Runway Gen-5 generates short video clips from a text prompt or a still image, with noticeably better physics than Gen-4: objects fall, fabric folds, and liquids pour without the melting look. It lives inside the Runway web editor and costs credits per second of output.
The basic plan includes 625 credits a month. A 5-second 1080p generation costs around 100 credits, so you get roughly six serious renders before you start paying.
What Is Runway Gen-5?
It is Runway’s current flagship video model, aimed at people making ads, music videos, and short narrative clips. You type a prompt, optionally attach a reference image, and the model renders a clip. There is also an image-to-video mode and a camera-control panel that exposes motion strength as a slider instead of prose.
Gen-5’s headline improvement is temporal consistency — the same character stays the same character across a handful of seconds. It is not perfect, but the gap to Gen-4 is obvious on faces.
How Do You Generate Your First Clip?
- Open the Runway editor and start a new project, then drag a blank asset onto the timeline.
- Choose Gen-5 in the model picker. Do not leave it on an older model by mistake; the picker remembers your last choice.
- Pick a ratio — 16:9, 9:16, or 1:1. Pick this before you prompt, because it changes framing significantly.
- Write the prompt. Subject, action, camera, light. Five seconds of video does not need more than about 25 words.
- Attach a reference image if you need a specific person or product. Runway Gen-5 holds identity much better with a reference than with description alone.
- Hit generate and wait. Expect 30-90 seconds per render at peak hours.
What Does Runway Gen-5 Cost?
| Plan | Monthly | Credits |
|---|---|---|
| Basic | Free | 625 |
| Standard | $15 | 625 + unlimited relaxed |
| Pro | $35 | 2,250 + relaxed |
| Unlimited | $95 | Relaxed unlimited |
Credits do not roll over. If you render in bursts, the Pro tier’s relaxed generations (slower queue, no credit cost) matter more than the raw credit count.
Where Does Gen-5 Still Fail?
Hands remain a problem, especially when they interact with objects. Fast lateral motion smears in the first few frames. Text on signs and shirts is still unreliable, and long dialogue with accurate lip sync needs the dedicated lip-sync pass, not the base model.
Reference images help identity but can flatten pose variety — if your reference is a straight-on portrait, most generations come back straight-on. Feed three-quarter angles when you want movement.
Who Should Use It?
Gen-5 suits short commercial work and social clips where a handful of strong seconds beats a slow edit. It is a poor fit if you need 30-second continuous shots or strict brand-accurate text on screen; those still belong in a traditional pipeline with rendered graphics.
Quick Summary
- Runway Gen-5 makes short clips from text or a reference image.
- Basic is free with 625 credits per month.
- A 5-second 1080p render costs roughly 100 credits.
- Reference images improve character consistency but can reduce pose variety.
- Hands, fast motion, and on-screen text remain weak spots.
