Generate high-quality miniature videos without VFX skills. Transform portraits into surreal glass ecosystem scenes perfect for social media and art projects.
Transform standard portraits into surreal, animated narratives where the subject appears miniaturized within a glass vessel. This image-to-video tool utilizes depth mapping and refraction simulation to render realistic glass textures and containment physics, creating a cohesive "world in a jar" aesthetic from a single static input.
Upload a clear portrait or full-body image of a person. Images with clean lighting and separation from the background produce the most realistic “human inside a bottle” effect.
The container style is automatically determined by the AI based on the input image and scene composition. The system intelligently applies a default transparent glass vessel that best supports the “human inside a bottle” illusion, ensuring optimal framing without manual adjustment.
Generate the video and preview the result. The output is a short MP4 clip where the subject appears to live and move inside the glass container.
The core of this tool is its ability to mimic the optical properties of real glass. The engine does not merely superimpose a bottle layer; it calculates how light bends (refraction) and reflects off the curved surfaces. This ensures the "person trapped inside a bottle" effect maintains visual coherence, treating the glass as a physical object rather than a flat filter.
To sell the illusion of a tiny human inside a large container, the AI automatically rescales the subject while generating a realistic depth of field. The subject remains in sharp focus, while the foreground glass and background elements receive a calculated bokeh effect. This mimics the focal length of macro photography, enhancing the sense of scale.
Static images are brought to life through subtle animation injection. The tool adds naturalistic movements, such as breathing, head tilting, or shifting weight, to the miniaturized subject. This motion synthesis is critical for converting a standard photo into a "living" video loop suitable for social media platforms.
Creators use this tool when they want to produce viral short videos for TikTok, Reels, or YouTube Shorts featuring a surreal “person trapped inside a bottle” effect. The output is designed to be visually striking within the first few seconds.
Artists and storytellers use the bottle containment effect to visualize fantasy concepts, dream sequences, or metaphor-driven narratives where a character exists inside a confined world.
Perfect for artists and roleplayers. Transform flat 2D drawings, anime sketches, or your beloved Original Characters (OCs) into volumetric 3D figures within a glass vessel. The AI interprets line work and shading to give your character depth, creating the illusion of a living digital figurine or a "trapped" 3D model inside the jar.
Our cloud-based architecture processes complex refraction physics in seconds, removing the need for local GPU power.
Access the tool seamlessly from desktop browsers or mobile devices, ensuring you can create content wherever you are.
Source images are processed securely and are not used to train public models without explicit user consent.
Images with a clean background and distinct lighting on the subject generally yield the most realistic composite. Avoid blurry or low-light photos.
Yes, videos generated on the Somake platform come with a commercial license, provided you own the rights to the original input image.
Short-form videos focused on surreal or metaphor-driven visual effects.
If the subject is too close to the edge of the frame in the original photo, the AI may struggle to "contain" them. Try using a photo with more headroom and space around the subject.