Imagen 4
Google's photorealism specialist. Three modes: Fast, Standard, Ultra.
Imagen 4 is Google's dedicated image generation model, built purely for photorealistic output. Three modes let you trade speed for fidelity: Fast for rapid iteration, Standard for balanced quality, and Ultra for maximum detail. Skin textures, fabric draping, material surfaces, and environmental lighting all render with a naturalism that other models struggle to match.
Example outputs
“A photorealistic portrait of a woman in her 30s, soft natural window light, shallow depth of field, editorial fashion photography”
“An architectural exterior render of a contemporary house, warm timber cladding, evening light, landscaped garden, photorealistic”
“A luxury watch on a slate surface, dramatic side lighting, macro product photography, visible dial details”
“A team of four people in a modern office, natural light, candid professional photography style, consistent characters”
“A ceramic vase collection on a linen backdrop, soft overhead light, editorial still life, earth tone palette”
“A sports car in a Mediterranean coastal setting, golden hour light, cinematic composition, photorealistic”
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 Imagen 4?
Jump into the Studio and start generating. Plans from £10/month.
Choose a PlanGoogle's flagship image model for photorealistic generation.
Imagen 4 is Google's highest quality image generation model. Built on the same foundation as the Gemini multimodal architecture, it delivers exceptional photorealistic rendering with strong prompt adherence and natural lighting. It is designed for developers and creators who need reliable, production-quality output at scale.
The multi-reference capability is a standout: Imagen 4 accepts up to 14 reference images in a single request, intelligently blending elements while preserving shape, geometry, and lighting consistency. This makes it practical for maintaining visual continuity across a project, from character sheets to product ranges to brand campaigns.
Character consistency works for up to 5 people across multiple generations without any fine-tuning required. Facial features, clothing, and distinctive traits carry across different scenes and compositions. Combined with batch processing for up to 4 variations per request, it enables efficient iteration on character-driven projects.
Photorealism and fidelity
Imagen 4 produces the most photorealistic output on the platform. Skin textures, fabric weaves, material surfaces, and environmental lighting all render with a level of naturalism that reduces the uncanny valley effect common in AI-generated imagery.
Character consistency
Maintain the same character across multiple generations: different poses, scenes, and lighting conditions. No LoRA training, no reference sheets. Describe the character once and Imagen 4 maintains recognition across outputs.
Multi-reference blending
Provide up to 14 reference images and Imagen 4 combines elements intelligently. Use this for maintaining brand consistency, combining style references, or building on previous generations.
Questions about Imagen 4.
Why Stensyl?
A small indie studio building creative tools the way they should be built. No VC theatre, no funnel games, no faceless support.