Runway Gen-4 Image
Cinematic stills from the team behind Gen-4.5 video. Reference-driven.
Runway Gen-4 Image brings the same visual DNA as their video models to still image generation. Built for consistency: upload up to 3 reference images and the model preserves identity, style, and lighting while transforming pose, angle, or environment. Strong prompt adherence, cinematic colour science, and an aesthetic range dial that lets you control how far the output deviates from your brief.
Example outputs
“A product photograph of a matte black wireless headphone on a marble surface, dramatic side lighting, warm studio background, premium product photography”
“A portrait of a female architect in a hard hat reviewing blueprints on a construction site, golden hour, shallow depth of field, editorial style”
“A cinematic wide shot of a modernist concrete house on a cliff overlooking the ocean at sunset, warm light, architectural photography”
“A character concept: a weathered sea captain in a wool coat, standing at a ship's wheel, stormy sky, cinematic lighting, concept art”
“A fashion flat lay of a curated outfit on a linen background: leather shoes, watch, folded shirt, minimalist styling, top-down product photography”
“A game environment concept: a neon-lit ramen shop in a narrow alleyway at night, rain, reflections, cyberpunk atmosphere”
How it works
Describe your vision
Type a detailed prompt or upload a reference sketch, photo, or mood board.
Choose your settings
Pick your resolution and aspect ratio. See the credit cost before you generate.
Generate in seconds
Your image is delivered in seconds. Download, iterate, or pipe into video.
Ready to create with Runway Gen-4 Image?
Jump into the Studio and start generating. Plans from £10/month.
Choose a PlanCinematic image generation with reference control.
Most image models treat each generation as independent. Runway Gen-4 Image is built around consistency. Upload a character reference and the model preserves that face across different poses, lighting conditions, and environments. Add a style reference and it locks the colour palette and visual treatment. Add a location reference and the architecture, materials, and spatial layout carry through. Up to three references at once, each doing a different job.
The aesthetic range control is worth understanding. Set it to 0 and the output stays as close to your prompt and references as possible. Push it to 5 and the model takes creative liberties with composition, colour grading, and stylistic interpretation. For client work where precision matters, keep it low. For concept exploration where surprise is welcome, push it higher. This single dial replaces a dozen prompt engineering hacks.
For design professionals, the reference system solves the biggest problem in AI image generation: maintaining a consistent visual identity across a set of images. Product designers can generate the same object in ten environments. Brand teams can produce a campaign where every image shares the same look. Character designers can test a face across costumes, ages, and lighting setups. Each generation takes seconds.
Up to 3 reference images
Each reference serves a purpose. One preserves character identity. Another locks visual style. A third anchors the environment. The model respects all three simultaneously, producing images that are consistent across a set without being repetitive.
Aesthetic range control
A single dial from 0 to 5 controls how much creative freedom the model takes. Low values produce precise, brief-accurate results. High values introduce stylistic variation and compositional surprise. No prompt engineering required.
Questions about Runway Gen-4 Image.
Why Stensyl?
A small indie studio building creative tools the way they should be built. No VC theatre, no funnel games, no faceless support.
Ready to create with Runway Gen-4 Image?
Professional image generation. Plans from £10/month.