AI assistants
The primary way to use Depicta. Connect Depicta to Claude on the web or Claude Desktop, or install a skill file in Claude Code, and your assistant can create visuals for you.
What is Depicta?
Depicta gives AI assistants and agents the ability to generate, edit, and process images. When your AI assistant is connected to Depicta, or has a Depicta skill file, it can produce images, icons, diagrams, and illustrated stories on your behalf.
How it works
Section titled “How it works”- You ask your assistant for a visual, or it decides on its own that one is needed — for example, when building a website or creating a presentation, it can generate images, icons, and diagrams as part of the workflow.
- Your assistant handles the rest — it writes the prompt, chooses a model (or uses the default for the task), and calls the Depicta API. Depicta provides guidelines and skill files that help your assistant make better decisions.
- You get the image — delivered as a signed URL (expires in 1 hour). Images are not stored permanently.
Why it works
Section titled “Why it works”An LLM is exceptionally good at writing image prompts. It has the full context of what you’re working on — the project, the style, the purpose — and can translate that into a detailed, specific prompt with all the small details you might forget or not think to include.
Depicta amplifies this in two ways. It helps your assistant write better prompts through curated guidelines. And it helps the image generation model better understand what’s needed through optimized prompt handling for each content type — images, graphics, and diagrams each get different treatment.
Three ways to use it
Section titled “Three ways to use it”CLI
Use the depicta CLI from your terminal or scripts. AI coding assistants like Claude Code use this directly.
API
Integrate image generation into your own apps and workflows via the REST API.
What can be produced
Section titled “What can be produced”Three content types, each with generate, edit, and combine operations:
| Content type | What it produces | Example |
|---|---|---|
| Image | Photos, illustrations, artistic renders | Product shots, concept art, backgrounds |
| Graphic | Icons, logos, clipart | App icons, UI elements, favicons |
| Diagram | Flowcharts, architecture, ER, sequence | System designs, org charts, timelines |
Plus stories — multi-scene illustrated narratives with consistent characters.
21 processing operations — crop, resize, remove background, blur, sharpen, convert formats, and more.
What Depicta provides
Section titled “What Depicta provides”- An API with endpoints for generating, editing, and combining images across multiple AI models
- 21 processing operations — your assistant can crop, resize, remove backgrounds, adjust colors, convert formats, composite images, and more
- Skill files that teach AI assistants how to use the API effectively
- Guidelines — a curated best-practices library that assistants query to write better prompts and choose appropriate models
- Multiple models with different strengths, costs, and capabilities. Each task has a default model, and your assistant can choose another.
- Content Credentials (C2PA) on every generated image, marking it as AI-generated, as the EU AI Act requires of us. If you publish realistic AI images, you may also need to label them; see Content Credentials.
Next steps
Section titled “Next steps” AI assistants Connect Claude, or set up a skill file.
CLI Install and generate your first image.
API Integrate image generation into your app.