Generate useful images and avatars for a real interface
Choose dimensions, styles, palettes and deterministic variations, then export editable SVG or raster PNG for your layout and test data.
Start with the space the image needs to fill
A placeholder becomes more useful when it resembles the shape and visual weight of the final asset. A wide hero image, a square profile picture and a small attachment thumbnail need different proportions. Set the destination width and height first, then choose a style that will still read clearly at that size.
The image generator offers landscape illustrations, gradients, geometric compositions, grids and solid backgrounds. These are original generated graphics, not photographs or remotely sourced stock images. You can build a social preview with a caption, a quiet background for a card or a predictable test image without leaving the tool.
Use a coordinated palette before adjusting individual colors
A preset palette supplies a background, foreground and accent that belong together. Sage, clay, ocean, lavender, ink and sunrise offer different visual directions while keeping the interface simple. A variation seed changes the composition without replacing the palette or dimensions.
Custom color controls accept six-digit hex values when you need to match an existing project. Switching back to palette mode restores the preset colors. Review captions against the actual background, particularly when you choose a low-contrast accent or plan to display the image smaller than its original dimensions.
Make the label fit the purpose
A title and subtitle are optional. They can identify an upload fixture, make a social preview readable or show which variation of an image is being reviewed. Text size adapts to its length and the canvas dimensions. Short captions work better for small cards; an image only thirty-two pixels wide cannot hold a readable paragraph.
The generated artwork does not fetch web fonts or remote images. SVG text uses a system font, and rasterized PNG captions can vary slightly between a browser and the hosted runtime. If a brand requires exact typography, export an editable SVG and apply the final font in your design workflow.
Choose between editable SVG and ready-to-use PNG
SVG stores the shapes, gradients and text as vector instructions. It stays editable and can scale to different display sizes without raster pixelation. PNG turns the composition into a fixed grid of pixels at the dimensions you requested. It is useful for profile-image uploads, social-image pipelines and applications that accept only raster files.
Both exports use the same composition settings. The browser renders PNG locally, while the hosted API uses the server renderer. Outputs are bounded to five MiB and sixteen million pixels. Large dimensions or complex artwork can reach those limits before a download is produced; reducing the canvas size is usually the simplest adjustment.
Use stable avatars for synthetic users
Profile fixtures should not require borrowing someone’s actual photograph. The avatar generator offers initials, symmetric identicons, orbital marks and a simple geometric robot. The name and seed determine the variation, and circle, rounded-square and square clipping make the result fit different interfaces.
Keep the same name, seed, style and palette when you want the avatar to stay stable across test runs. Change the seed for a different visual without changing the identity label. Identicons are decorative here: they do not authenticate a person, verify a key or replace an account’s security checks.
Make image generation reproducible from a client
The same generator is available through the web, HTTP API, SDK, CLI, SSH and MCP. Pass a title or name as input, then supply options such as style, theme, seed, width, height and format. The per-tool descriptor includes option names, choices and defaults, so an integration does not have to infer them from a screenshot.
For a development dataset, save the generation settings with each fixture. This makes a visual regression easier to reproduce and avoids a dependency on a changing remote image URL. The SVG is deterministic for a fixed recipe. PNG rasterization can differ in font details across environments, so compare the intended renderer when pixel-for-pixel screenshots are important.
socket-help describe test-media-placeholder-images
socket-help run test-media-placeholder-images \
--options '{"style":"landscape","theme":"ocean","width":"1200","height":"630","format":"png","seed":"41"}' \
< title.txt
# Decode the returned file artifact’s base64 field to save the PNG.