Reve is now available on fal

Reve is now available on fal

We are excited to announce the release of Reve now available from day 0 on fal. Reve is a state-of-the-art image generation model that brings visual intelligence, layout design, and editing into one seamless creative workflow.

With text-to-image, image editing, and multi-reference generation, Reve gives creators precise control over composition, consistency, and storytelling. Whether you’re designing characters, products, or full visual narratives, Reve turns imagination into polished, spatially coherent imagery

Understanding Reve's Core Strengths

Visual Intelligence

Reve demonstrates a level of visual intelligence that goes far beyond aesthetic style. It understands how objects relate to each other in space, how a design should stay consistent across multiple views, and how visual information should be organized on a page. In architectural compositions, it can generate complete design layouts; elevations, sections, perspective views, and annotations, all aligned in proportion and lighting. The text within these designs integrates naturally, appearing in correct perspective and positioned exactly where technical labels or callouts would be placed.

Link to generation: https://fal.ai/models/fal-ai/reve/text-to-image?share=21608916-5320-4369-a1bc-78b8a096f684

For narrative layout, the model composes full cartoon storybooks with panel grids, safe margins, and a clear reading order. Characters remain consistent in color palette, facial features, and scale from panel to panel. Speech balloons avoid occluding faces, tails point to the speaker, and background elements carry forward so continuity is visible from the first frame to the last. The result is pages that are ready for print without manual reflow.

Link to generation: https://fal.ai/models/fal-ai/reve/text-to-image?share=c3efc4cd-25c9-460e-83c2-fdada011471a

Image Editing

Reve excels at image editing, allowing creators to modify or expand visuals while preserving the integrity of the original design. It can replace elements, adjust lighting, or alter composition without breaking perspective, texture, or style. When editing products or characters, Reve maintains consistency in proportions, materials, and pose, so the final result feels like a continuation, not a patch.

The model’s spatial and visual understanding ensures that new additions blend seamlessly with existing content. Shadows fall correctly, reflections align with light sources, and background depth remains intact. This makes Reve ideal for iterative workflows: updating product designs, refining ad scenes, or generating new variations from a single concept, all while keeping the overall aesthetic coherent and believable.

Input

Input Image

Output

Output Image

Prompt: Make him blonde and add cool sunglasses

Reference to Image

Reve lets you direct your scene with precision by using multiple reference images as visual ingredients. You can guide composition, lighting, character design, or material style simply by providing reference images, each reference contributes to the final look in a controlled and interpretable way. This means you can define exactly which elements to keep consistent, such as a character’s face, a product’s texture, or the atmosphere of a setting.

You can combine several references in a single generation to build complex scenes: one image for environment, another for pose, another for color palette or costume. Reve interprets these inputs intelligently, merging them into a cohesive image that respects spatial relationships, proportions, and lighting direction. The result feels composed, not collaged; a fully realized scene directed by you.

Inputs

Input 1 Input 2

Output

Output

Prompt: This woman sitting on this bed and knitting, close shot image

Unparalleled Realism

Reve produces images with striking realism, blending photographic detail with physical accuracy. Materials respond correctly to light, Edges stay crisp where they should, but soften in areas where focus or depth of field come into play, creating a cinematic sense of lens behavior.

This realism extends to composition and perspective. Reve understands how scale, lighting direction, and atmospheric depth interact, so scenes feel grounded and cohesive. Whether it’s a product photo, a landscape, or a character portrait, every element aligns in geometry and illumination, giving the image a convincing sense of presence.

A domestic cat curiously approaches a majestic snow tiger in a serene, snow-covered landscape. The unlikely pair tentatively interacts, showcasing a moment of gentle curiosity and surprising connection amidst the winter's stillness. It is a simple meeting between two distinct creatures in a shared environment.
Engines roar as sleek racing cars line up at the starting grid, ready to unleash their power. The scene follows them speeding around the track, highlighting their velocity and design as they compete for the lead. Ultimately it culminates in the checkered flag waving, symbolizing victory.

Getting Started with Reve

The easiest way to explore Reve's capabilities is through Fal's Playground, where you can experiment with prompts and see immediate results. A detailed guide on how to integrate Reve into your platform is available in our API documentation.

Endpoints:

Reve | Text to Image | fal.ai
Reve’s text-to-image model generates detailed visual output that closely follow your instructions, with strong aesthetic quality and accurate text rendering.
Reve | Image to Image | fal.ai
Reve’s remix model lets you upload an reference images and then combine/transform them via a text prompt
Reve | Image to Image | fal.ai
Reve’s edit model lets you upload an existing image and then transform it via a text prompt

Stay tuned to our RedditblogTwitter, or Discord for the latest updates on generative media and the new model releases!