Open-source AI image and design agents (2026)
AI image and design agents with public source code you can read, self-host and extend. Here are the top 25 of 28 from the AgentScout directory.
Quick answer: the top picks on this list, ranked by GitHub stars and listing maturity, are ComfyUI, Diagram Design and Penpot, out of 25 tools listed below. At a glance: 25 open-source, 25 free or freemium and 20 local or self-hosted across 25 link-checked tools (last verified ).
Verified on : every tool link on this page was checked by our daily link check. No paid placements.
Ranked comparison (top 10)
| # | Tool | Pricing | Best for | Open source | Runs locally | GitHub stars |
|---|---|---|---|---|---|---|
| 1 | Open source + cloud | generate video, generate images | Yes | 143,460 | ||
| 2 | Open source + cloud | build an agent, make diagrams | Yes | Self-hosted | 62,392 | |
| 3 | Open source | UI design | Yes | 61,747 | ||
| 4 | Open source | control a computer, connect tools via MCP | Yes | 36,437 | ||
| 5 | Open source | build apps without code, write code | Yes | 27,013 | ||
| 6 | Open source + cloud | build an agent, connect tools via MCP | Yes | Self-hosted | 19,762 | |
| 7 | Open source | write code, connect tools via MCP | Yes | Self-hosted | 16,077 | |
| 8 | Open source | build an agent, automate workflows | Yes | Self-hosted | 11,861 | |
| 9 | Open source | write code | Yes | Self-hosted | 11,652 | |
| 10 | Open source + cloud | connect tools via MCP | Yes | Self-hosted | 8,954 |
Updated
How we picked and ranked: tools are matched by category, the pricing each maker states and what each tool says it does. They are ranked by GitHub stars as recorded on each listing (link checks 2026-10-28); tools without a tracked repo follow, by listing maturity. No paid placements.

1. ComfyUI
Node-based open-source engine for image and video generation workflows. Run models like Wan, LTX and Hunyuan locally or in Comfy Cloud, with an agent that builds workflows.
Pricing (as stated): Desktop app free and open source (GPL-3.0); Comfy Cloud Standard $20/mo ($16 yearly), Creator $35/mo, Pro $100/mo, Team $700/mo · Active

2. Diagram Design
Editorial diagram design for Claude Code, Codex, GitHub Copilot, Cursor, Factory Droid, and Pi. 44 diagram types.
Pricing (as stated): Open source + cloud · Mature

3. Penpot
Open-source design and prototyping tool with an official MCP server so AI agents can read and edit design files.
Pricing (as stated): Free and open source (MPL-2.0); cloud Unlimited $7/editor/mo (capped at $175), Enterprise $25/member/mo · Active

4. Next AI Draw.io
Create and edit draw.io diagrams by chatting with AI, as a web app, desktop app or MCP server.
Pricing (as stated): Open source · Mature

5. Onlook
An open-source visual editor for React and Tailwind apps where you design on a canvas and an AI agent writes the code. Edit layouts like in Figma, then export clean code to your own repo.
Pricing (as stated): Free and open source (Apache-2.0); hosted cloud version in early access. · Mature

6. Toonflow App
、 Toonflow: AI , 、 、 、 。 、AI Agent , 、 、 、MCP 。Open-source AI short drama and motion comic creation tool with an infinite canvas, AI agents, storyboarding, image and video generation. Self-hosted and.
Pricing (as stated): Open source + cloud · Mature

7. Framelink Figma MCP
MCP server that gives Cursor and other coding agents Figma layout data so they can implement designs accurately in one shot.
Pricing (as stated): Open source · Mature

8. Fireworks Tech Graph
Generate production-quality SVG+PNG technical diagrams from natural language. 7 styles, UML support, and AI/Agent workflow patterns
Pricing (as stated): Open source · Mature

9. Loop Engineering
Practical patterns, starters & CLI tools for loop engineering with AI coding agents. Design systems that prompt and orchestrate agents (inspired by Addy Osmani and Boris Cherny).
Pricing (as stated): Open source · Mature

10. Dashboard Icons
3,400+ curated icons and logos for dashboards and AI agents. SVG, PNG and WEBP on a free CDN, plus an MCP server for Claude Code, Codex, ChatGPT, Cursor, Hermes and OpenClaw
Pricing (as stated): Open source + cloud · Mature

11. Anti Slop
Rules for an AI coding agent to filter out generic AI-generated UI designs, text, and code
Pricing (as stated): Open source · Mature

12. Talk to Figma MCP
MCP integration that lets coding agents read and edit Figma designs.
Pricing (as stated): Open source · Mature

13. draw.io MCP
The official draw.io MCP server, so Claude, ChatGPT, Cursor and other assistants can create and edit flowcharts, architecture diagrams and other draw.io diagrams right in the chat. Free and needs no sign-in.
Pricing (as stated): Free; hosted MCP App server needs no sign-in. · Mature

14. OpenKnowledge
OpenKnowledge is a free, local-first markdown editor and knowledge base built for working alongside AI agents. It looks and feels like a Notion-style WYSIWYG editor, with callouts, tabs, Mermaid diagrams and media blocks, but every page stays a plain markdown file on your disk. Coding agents such as Claude Code, Codex, Cursor, Gemini CLI and dozens of others can read and write the same files through its integrations, so specs, team wikis and LLM wikis stay in one place that both people and agents keep up to date. It ships as a desktop app plus a web UI and CLI.
Pricing (as stated): Free and open source (GPL-3.0): desktop app, web UI and CLI · Active

15. Design Extract
Extract any website's complete design system with one command. DTCG tokens, semantic+primitive+composite, MCP server for Claude Code/Cursor/Windsurf, multi-platform emitters (iOS SwiftUI, Android.
Pricing (as stated): Open source + cloud · Active

16. shadcn/lint
Linter for Tailwind design systems from shadcn that turns your design rules into checks coding agents can run and fix against.
Pricing (as stated): Open source · Active

17. CodeBoarding
Generates interactive architecture diagrams of codebases.
Pricing (as stated): Open source · Active

18. Figma Console MCP
Design-system-focused MCP server that reads and writes Figma and checks design-code parity and accessibility.
Pricing (as stated): Open source · Active

19. Oil UI
UI design skill for Claude Code, Codex and other coding agents that explores several distinct visual directions, compares them side by side and refines the chosen one against real screenshots. Free open-source skill.
Pricing (as stated): Open source · Active

20. PR Lens
PR Lens helps people review code that agents wrote. For each pull request it draws animated architecture and data-flow diagrams inside the PR, showing which parts of the app were added, changed or removed, what else depends on the change and how a request now moves through the system. It also flags likely bugs, security holes and design problems in the same review comment, with a ready prompt to hand each fix to your coding agent or an autofix. The open-source core runs locally and installs as an agent skill.
Pricing (as stated): Open-source core (MIT), runs locally; installs as an agent skill · Active

21. Py Gpt
Desktop AI Assistant powered by GPT, Gemini, Claude, Grok, Ollama, DeepSeek, Perplexity, and more - chat, agents, tools, MCP, plugins, RAG, vision, voice, image and video generation, speech synthesis.
Pricing (as stated): Open source + cloud · Active

22. Erd Editor
Design database schemas as ERDs in your browser, VS Code, JetBrains IDEs or Obsidian. Import SQL, DBML and GraphQL, export DDL and code, and let coding agents edit live over MCP
Pricing (as stated): Open source + cloud · Active

23. Terravision
Cloud architecture diagrams in both directions: AI prompt or JSON to diagram, diagram to Terraform via MCP server, and Terraform to diagram via CLI or CI/CD. Using Official AWS, Azure and GCP icons
Pricing (as stated): Open source · Active

24. Pixeltable
Multimodal data layer that combines storage, transforms, indexing and retrieval in one Python library for building agents over images, video and documents.
Pricing (as stated): Open source + cloud · Active

25. MiniMax MCP
Official MiniMax MCP server that lets agents call MiniMax text-to-speech, voice cloning, image and video generation.
Pricing (as stated): Open source · Active
FAQ
What are the best open-source AI image and design agents?
Ranked by tracked GitHub stars, the top open-source AI image and design agents in our directory are ComfyUI, Diagram Design, Penpot, Next AI Draw.io, Onlook. Stars measure popularity, not fit, so compare a few against your own task.
How many open-source AI image and design agents does AgentScout list?
28 tools match. This page shows the top 25. Pricing mix on this page: Open source (17), Open source + cloud (8).
How much do these tools cost?
Pricing is what each maker states on their site. Open-source tools are free to self-host but may need a paid model API key or offer paid cloud plans; freemium tools have a free plan with limits. Check each official site.
Where can I see every tool for images and design?
Our task page lists all 59 tools we match to images and design.
Related: All tools for images and design · MCP servers (7) · coding (7) · productivity (3) · frameworks (3)
By role: For marketers · For designers
All popular searches · AI agents by role · Free tools
Best-of lists: Cursor alternatives · All best lists
Related categories: Coding · MCP servers · Frameworks · Productivity
Facts come from our directory listings; pricing is as each maker states it and can change. Check official sites before you choose.