What can I automate with the Colormind integration?

Generate color palettes with Colormind, the free keyless color scheme generator driven by a deep learning model trained on photographs, film stills and popular art. Get a five-color palette from nothing, build one around the colors you already have, ask what goes with a single color, or get a UI palette with a light shade, two accents, the main brand color and a dark shade built around your brand color. Draw from the general default model, the ui model, or one of the models that rotate in daily, trained on a film, a game or a style of photography, and check which are loaded before a step relies on one. Lodol supports 8 actions for Colormind, which you can combine with any other app in a workflow.

How do I connect Colormind to Lodol?

Connect Colormind once in your workspace and every action and trigger it supports can be added to a workflow, by you in the editor or by your AI.

All integrationsImages & Design

Colormind

Generate color palettes with Colormind, the free keyless color scheme generator driven by a deep learning model trained on photographs, film stills and popular art.

Start Free Browse all integrations

actions: 8

Actions

Steps a workflow can run in Colormind.

Generate a color paletteGenerate a five-color palette from nothing with one of Colormind's models: the general default model, the ui model for interfaces, or one of the models that rotate in daily, trained on a film, a game or a style of photography. Every color comes back as a hex code, a CSS-ready one and RGB, and a ui palette labels each slot with its role.
Generate several palettesGenerate up to ten palettes with one model in a single step, to pick from or to post as options. Each draw is random, so the palettes differ, and each costs Colormind one request. Every palette comes back with its colors, its hex codes and its CSS-ready ones.
Complete a paletteBuild a palette around the colors you already have. Give up to four colors, in the slots you want them in, and Colormind fills the rest to go with them. Each color that comes back says whether its slot was locked and, because the model nudges a locked color by a small amount to fit, what was asked for and whether it changed, so a brand color that must stay exact can be checked or put back by the next step.
Suggest matching colorsGet four colors that go with one: the color is locked into one slot of a palette and Colormind fills the other four around it. Choose the slot when it matters; with the ui model, slot 3 is the main brand color and the rest become its shades and accents. The four suggestions come back on their own, with the whole palette beside them, and the answer says whether the model nudged the color you gave.
Generate a UI paletteGenerate a palette for a user interface with Colormind's ui model, built around your brand color when you give one. The five colors come back by role as well as in order: a light shade for backgrounds, a light accent, the main brand color, a dark accent and a dark shade for text, so a step can read light_shade or dark_shade by name. The answer says whether the model nudged the brand color to fit.
List modelsList the models Colormind has loaded today. default and ui are always there; the others are trained on a film, a game or a style of photography and are replaced every day at 08:00 UTC, so the answer says which are permanent, which are today's, and when the next change is.
Check a model is availableCheck whether a model is loaded today before a step relies on it. Models other than default and ui change every day, so a workflow built on one should ask first. A model that is not there is reported as a value rather than a failed step, with the nearest names that are, so the workflow can fall back to one of them.
Check service statusCheck whether Colormind is answering, how quickly, and how many models it has loaded. This is the one action that reports a problem as a value instead of failing the step, so a workflow can decide what to do about it. It also says whether the check fell in the daily reload window at 08:00 UTC, when the service is down for about 30 seconds while the day's models load, and when the next reload is.