AI Tool

Doop Review

Doop is an open-source, multiplayer design canvas where humans and AI agents collaborate live to create and revise web designs.

shipped Sep 2, 2026agentsfreemium

Monthly visits38/mo

agentsimage-generation

Why it matters

1Launched on Product Hunt on September 2, 2026.

2Supports AI agents like Claude Code, OpenAI Codex, and OpenAI GPT-5.6-sol.

3Offers a freemium model with a Hobby tier at $0 and a Garden tier at $6/month.

4Generates HTML-based 'Frames' suitable for web UI, landing pages, and product screens.

About Doop

Business Model

Freemium SaaS

Usage Pricing

$6 per plant

Platforms

Web, macOS

Target Audience

Designers and teams looking to enhance collaboration with AI agents.

Pricing Plans

Hobby

• 3 plants
• basic schedule

Garden

$6/mo

• Unlimited
• sensor sync

Cost Examples

• First 3 plants are free, each additional plant costs $6/month

API Docs GitHubOpen Source

Specs

API Available

Yes, public API

overview

What is Doop?

Doop is a collaborative AI design canvas tool developed by Doop that enables designers and creative teams to create and revise web designs alongside AI agents. It provides an infinite canvas where AI agents join through MCP, sketching frames, filling real copy, and reviewing their own work while users steer.

features

Key Features of Doop

Doop integrates advanced AI capabilities with a collaborative design environment, focusing on real-time interaction and HTML-based output.

Live AI Design Streaming: Observe AI agents progressively building HTML Frames on the canvas in real-time.
MCP Agent Support: Compatibility with various AI agents including Claude Code, Codex, Cursor, Claude Desktop, and Copilot.
Built-in Doop Agent: Assign design tasks to an integrated agent without requiring an external MCP session.
Rendered Output Inspection & Self-Review: Agents utilize a headless renderer to generate screenshots for self-correction of visual issues.
Design Memory and Taste Profile: User feedback is distilled into a 'taste profile' to propose and apply consistent style rules across designs.
True Multiplayer Collaboration: Supports live cursors, presence indicators, per-frame editing, undo/redo, and element-pinned comments.
HTML-based Frames: Designs are rendered as actual HTML within sandboxed iframes, facilitating direct web development use.
Feedback as Tasks: Comments and replies can be converted into actionable tasks for AI agents to claim and execute.
Function Calling: Proprietary function calling capabilities for AI agents.

use cases

Who Should Use Doop?

Doop is designed for designers and creative teams seeking to integrate AI agents directly into their design workflow, particularly for web-centric projects.

Designers and Creative Teams: For collaborative design between humans and AI agents on web UI, landing pages, and product screens.
Web Developers: For generating HTML-based designs that are directly usable for web development.
Teams Requiring Iterative Design: For workflows where humans sketch initial concepts and AI agents refine them.
Users Seeking Live AI Collaboration: For observing AI agent design work in real-time and providing structured feedback mid-flight.

how to use

How to Use Doop

To begin using Doop, users can access the platform via web or macOS and connect their existing AI subscriptions for agent functionality.

1Access Doop: Navigate to doop.design or download the macOS application.
2Connect AI Subscriptions: Link existing Claude or ChatGPT accounts to enable full agent capabilities.
3Utilize Doop Agent: Assign design tasks to the built-in Doop Agent for initial free tasks, then via connected subscriptions.
4Collaborate on Canvas: Begin designing on the infinite canvas, sketching frames and providing feedback to AI agents.
5Review and Refine: Observe AI agents' real-time design streaming and leverage their self-review capabilities for iterative refinement.

pricing

Doop Pricing & Plans

Doop operates on a freemium model, offering a free tier and a paid subscription. Users are expected to provide their own AI subscriptions for most agent functionalities.

Hobby: $0 per month. Includes a limited number of free tasks with the built-in Doop Agent. Subsequent tasks run on the user's connected ChatGPT subscription or OpenAI key.
Garden: $6 per month. This tier is listed with a usage pricing of $6 per plant, with the first 3 plants free and each additional plant costing $6/month. Connecting a user's own MCP agent directly to the canvas is not metered by Doop.

Enjoying this? Get one like it in your inbox each morning.

one email a day · unsubscribe in two clicks · no third-party tracking

Pros

+Enables real-time, live collaboration between human designers and AI agents on a shared canvas.
+Generates actual HTML-based 'Frames' directly usable for web development, unlike mock-up tools.
+AI agents can inspect rendered output via screenshots and self-review their designs before completion.
+Supports a Multi-Agent Communication Protocol (MCP) for integrating various AI models like Claude Code and OpenAI Codex.
+Features design memory and a 'taste profile' to ensure consistency and apply durable style rules across projects.
+Offers a free tier ('Hobby') allowing users to explore core functionalities.

Cons

−Requires users to 'bring their own AI subscriptions' (e.g., ChatGPT, Claude) for full agent functionality beyond initial free tasks.
−As of August 2026, user reception indicates it is 'a bit too early and not ready to replace Figma anytime soon,' suggesting maturity limitations.
−The 'Garden' pricing tier includes usage-based costs per 'plant,' which may not be immediately intuitive for design software.
−Focuses specifically on HTML-based web design, potentially limiting its utility for other design disciplines like native app UI or print.

Similar Tools

Doop vs Competitors

Doop differentiates itself by treating AI agents as first-class collaborators on a shared canvas, focusing on live, HTML-based design generation.

tldraw↗

It's an open-source infinite canvas with an AI feature ('Make Real') that transforms hand-drawn sketches into editable UI components.

While tldraw offers an infinite canvas and AI-powered transformation of sketches, it doesn't explicitly feature 'AI agents designing beside you' or a multi-agent collaboration protocol like Doop. The AI interaction is more on-demand for specific transformations rather than continuous co-design.

Magician (Figma Plugin)On Stork

It integrates AI capabilities like text generation, icon generation, and image generation directly into Figma's infinite canvas, assisting designers with content and assets.

Magician provides AI assistance within Figma's design environment, offering tools for content and asset creation. However, it doesn't replicate Doop's concept of autonomous AI agents actively participating in the design process, sketching frames or reviewing work.

Miro AI↗

Miro AI offers features like idea generation, clustering, and summarization directly on its infinite collaborative whiteboard, helping teams brainstorm and organize thoughts.

Miro AI focuses on enhancing brainstorming and ideation on a collaborative canvas with AI-powered organization and generation. While it offers AI assistance on an infinite canvas, it doesn't provide the 'AI agents designing beside you' functionality for visual design elements like Doop.

UizardOn Stork

Uizard uses AI to rapidly convert text descriptions, hand-drawn sketches, or screenshots into editable wireframes and prototypes on a design canvas.

Uizard excels at quickly generating initial UI designs from various inputs, offering a canvas for editing. While it uses AI for design generation, it doesn't feature the continuous, agent-based co-design and review process that Doop emphasizes.

See every Doop alternative, compared→