Gen-4 Video
Gen-4 Video creates motion from an image and a text prompt. Runway’s research introduced the family around visual consistency across characters, locations and objects. It is an earlier generation than Gen-4.5.
Establish the look before generating movement
Runway describes using visual references to preserve subjects and style across new views. That makes the starting image an important part of a shot’s design. The family’s consistency claims describe intended behavior, not a guarantee that every generated frame will preserve every detail.Introducing Runway Gen-4
Distinguish the product from the API model
The creative guide still lists Gen-4 and Gen-4 Turbo in the model selector. The current Dev catalog lists Turbo as gen4_turbo. This guide does not assign the non-Turbo product a guessed API identifier.Creating with Gen-4 VideoRunway Dev model catalog
Practical guidance. When continuing an older project, save its reference image and compare the new clip against that source. Examine the beginning and end as well as the middle, where changes to faces, objects or materials can be easier to miss.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Provider product |
| Required input | A starting image plus a text prompt.Creating with Gen-4 Video |
| Clip duration | 5 or 10 seconds at 24 fps.Creating with Gen-4 Video |
| Product price | 12 credits per second, according to the Gen-4 guide.Creating with Gen-4 Video |
Related models
A video-generation model that accepts text prompts or starting images.
Launch date not confirmedAPI · Provider product- Gen-4 Image
Runway · Gen Image
Runway’s image model with visual references, composition controls and 720p or 1080p output.
Launch date not confirmedAPI · Provider product - Gen-4 Turbo
Runway · Gen
Runway’s faster Gen-4 image-to-video model, with API access and a lower listed price than Gen-4 Video.
Launch date not confirmedAPI · Provider product