Ardot is an AI-driven intelligent tool launched by Tencent. It provides AI-assisted design, vector editing, MCP automation and team collaboration capabilities, covering the entire process from prototype to delivery.

Ardot - Tencent AI-driven one-stop intelligent design platform

Ardot’s core parameters and statistics

Ardot is not a simple "AI poster generator". It is an AI-driven intelligent design platform launched by Tencent. It is positioned against professional design tools such as Figma, but uses AI capabilities throughout the entire design process - from prototype, vision, collaboration to development and delivery. The core difference is that generative AI is directly embedded into professional vector editing contexts, rather than just lightweight template splicing.

Projects	Public Information
Official positioning	AI-driven intelligent design tool/AI design agent collaboration platform
Producer	Tencent (Tencent Cloud)
Product form	Web online + macOS desktop client (Windows desktop client in internal testing)
Core Competencies	AI text/pictures for UI, AI text/pictures for pictures, vector editing, dynamic layout, component system MCP automation, team collaboration
Target scenario	UI/UX design, marketing vision, design system management, industry-device-research collaboration, design draft conversion to code
Target users	Product managers, designers, R&D engineers, operations, small and medium-sized enterprise design teams
Supported platforms	Web, macOS (iOS/Android mobile terminal viewing and commenting)
Home	China (Tencent Cloud)
Billing form	Starter free version + Pro/Team/Enterprise subscription system, currently Pro/Team is free for a limited time
Pricing system	AI Credits measurement system (unified measurement unit for AI capabilities)
MCP capabilities	Supports Model Context Protocol and provides 18 design operation tools
Figma migration	Supports direct import of .fig files, fully retaining layout, style and components

Product Boundary: Ardot is an "AI design agent collaboration platform" for professional industry, design and research teams. It is not a simple AI poster generation tool. It is suitable for scenarios that require complete design system management, multi-person collaboration, and development delivery; it is not suitable for non-design scenarios that only require the generation of a single image at one time.

Capability source: Relying on Tencent's technology accumulation in AI large models (Hunyuan, etc.), cloud computing and social ecology, generative AI is deeply embedded into the workflow of professional design tools.

Users and market recognition of Ardot

As a product of Tencent Cloud, Ardot naturally enjoys the advantage of access to Tencent’s ecosystem. Its public version number has been iterated to v1.51 (2026-07-01), and it has maintained an almost daily/weekly continuous update rhythm since its launch in 2025, indicating that Tencent has invested continuous R&D resources in this product.

Market Positioning: Ardot’s direct competitors include professional design tools such as Figma (collaborative design), MasterGo (domestic alternative), instant design, and Canva (lightweight design). Its differentiation lies in "AI native" - ​​not plug-in AI functions on design tools, but using AI from the bottom up as the core interaction method of the design process.

True value: The core pain point it hits is "the multi-role collaboration efficiency of the production, design and research team". Product managers use AI to quickly produce prototypes, designers use AI to assist in visual refinement, and R&D engineers directly read design files through MCP to generate code - this saves a lot of communication and rework time compared with the traditional serial process of "designing and then delivering for development." However, as a relatively new product, its market share and maturity are still on the rise, and the official has not disclosed the specific user scale and number of enterprise customers.

Ardot’s cost advantage

• C-side/Individual: Usually a free version is provided to experience the core functions, and high-frequency use requires a paid package subscription.
• API/Developer: Billed by call volume, suitable for development teams that can be flexibly integrated into their own systems.
• Enterprise/Privatized: Contact the business owner for customized quotation and deployment plan. The specific price is subject to the official real-time pricing page.

Main functions of Ardot

Ardot's functions cover the complete design process from "idea to code" and are divided into four capability layers:

AI intelligent design layer

• Text/Illustrated UI design draft: Use natural language to describe the interface requirements, or upload reference images, AI will generate the first draft of the UI in real time, and the generated results will be based on the team component library to maintain brand consistency.
• Text/Illustration: Enter a text description or upload a reference image, and AI will generate visual materials such as illustrations, pictures, and operational posters, supporting multiple style selections and secondary adjustments.
• Batch generate vector icons of the same style: Enter text or sample images, and generate a complete set of vector icons with a unified style with one click, which can be saved as a template for reuse.
• Intelligent Frame Selection Modification: Frame select any area of ​​the canvas and use natural language to refine local content without manually adjusting layers.
• AI Assistant: An intelligent assistant that runs through the entire design process, supporting advanced capabilities such as intention clarification, customized file upload analysis, and specified Skill upload.
• AI thinking process visualization: Displays the reasoning chain in the AI ​​generation process, improving the transparency and controllability of intelligent interaction.

Professional vector editing layers

• Vector Editing: Supports vector grid, Boolean operation 3D rotation, and enhanced shape tools (bevel, polygon, etc.) to meet the creative needs from simple icons to complex illustrations.
• Dynamic Layout System: Grid layout, percentage positioning, and true proportion locking ensure that the design draft is perfectly presented in different screen sizes and reduce development adaptation communication.
• Auto Layout: Supports basic and advanced automatic layout, including complex nesting, responsive adjustment of Canvas Stacking and custom stacking order.
• Component system: main component + instance mechanism, component variants (Variants), component properties (Component Properties), and supports publishing team component libraries.
• Design Variable System: Design Tokens such as colors, values, and strings, supporting automatic switching of multiple themes such as Light/Dark, and variable linkage global modification.
• Style Management: Fill (solid color/gradient/picture), stroke, shadow, blur and other effects, color/text/effect styles can be reused.

Collaboration and Management

• Multi-person real-time collaboration: Supports multiple people to edit the same file online at the same time, and view member cursor positions and operations in real-time.
• Comments and Annotations: Supports direct marking of comments, @members, task assignments on the canvas, and review comments will be left traced throughout the process.
• Intelligent Permission Center: Super management and fine allocation of permissions, supporting viewing/editing/management and other levels, making key resources safe and controllable.
• Comprehensive behavior traceability: Operation logs are fully recorded, and asset changes are clearly traceable to meet auditing needs.
• Seamless handover mechanism: Automatic recovery of permissions upon resignation, automatic transfer of design assets, and complete retention of historical versions.
• Development collaboration (R&D mode): visual annotation, component variant viewing, one-click copying of multi-end code (React/Vue/mini program/iOS/Android), and synchronous export of motion parameters.

MCP Automation and Development Delivery Layer

• MCP Server (Model Context Protocol): Allows AI Agent to directly read and operate Ardot design files to achieve two-way linkage between design and code. Provides 18 tools (continuously updated):
  • Query class: fetch_editor_state, fetch_file_info, batch_read, capture_screenshot, capture_layout
  • Design system: fetch_variables, apply_variables, fetch_component_lib, fetch_guidelines, search_style_guide, build_style_guide
  • Edit operations: batch_edit (6 operation types), locate_available_space, create_new_page
  • Resources and Exports: scan_exportable_resources, export_nodes, upload_images
  • Accessibility: get_available_fonts
• Direct code restoration through CodeBuddy: The design draft is sent to CodeBuddy with one click through the MCP protocol, and high-reduction front-end code (React, Vue, applets, etc.) is automatically generated.
• Custom Skill: Precipitate team design specifications and workflow, allowing AI to generate designs according to team standards.

Ardot’s model and version evolution

Ardot is divided into desktop client (macOS/coming soon to Windows) and Web online version, and the functions of both are synchronized. The following is the key version context compiled based on the official version release log:

Version number	Date	Core changes
v1.51	2026-07-01	GPU rendering rollback, large file opening optimization, points increase package signboard AI assistant lag repair
v1.50	2026-06-30	MCP connection stability optimization, CDN storage strategy optimization, multiple rendering defect repairs
v1.49	2026-06-26	Mini program collaboration entrance MCP image upload/download, desktop performance optimization
v1.48.194	2026-06-22	Dark mode is fully online, account merging login, variable table search and filtering
v1.45.182	2026-06-09	AI thinking process visualization AI assistant multi-language, noise effect
v1.43.172	2026-06-03	AI intent question clarification, customized file upload analysis, specified Skill upload
v1.37.148	2026-05-21	Windows-side functions align with macOS and AI image editing capabilities
v1.34.137	2026-05-18	Smart box selection to modify AI style guide retrieval framework Slide node type
v1.31.127	2026-05-13	AI image modification and UI partial redrawing Canvas Stacking layout
v1.28.118	2026-05-10	Canvas paged loading, .ardot file import, full access to domestic large models
v1.26.114	2026-05-06	AI generated slide PDF export
v1.19.91	2026-04-21	AI assistant newly launched (batch SVG icons, Vincent diagrams, local component library UI generation)
v1.13.62	2026-04-03	Desktop supports pasting pictures MCP setting entrance
First version	~2025	Ardot goes online, providing AI-assisted visual creation and vector editing capabilities

Version Rhythm: Starting from v1.13 (2026-04), Ardot has entered an intensive iteration period, with versions released almost every day. The launch of v1.19’s AI assistant is a functional milestone. This reflects that the product is in a period of rapid growth, with functionality and stability advancing simultaneously.

Technical advantages of Ardot

Mechanism: Ardot adopts an "AI native" design architecture instead of plug-in AI functions on traditional design tools. AI capabilities run through the entire link of design document creation, editing, collaboration, and delivery, and realize two-way communication with external AI Agents through the MCP protocol.

Effect: Technical advantages at three levels:

1. Generation efficiency: Text/pictures directly generate UI design drafts and visual materials, compressing the drafting time from "requirements to design drafts" from hours to minutes. Capabilities such as batch generation of vector icons of the same style and smart frame selection and modification further reduce duplication of work.
2. Design consistency: AI generation is based on the team's component library and design variables to ensure that the output conforms to brand specifications and avoids the "orphan component" problem. The design variable system supports automatic switching of multiple themes and linked global modification of variables, making it suitable for large-scale design system management.
3. Delivery is guaranteed: Through MCP Server + CodeBuddy direct code generation, the restoration degree of design draft to code can reach more than 99%. Annotation, code copying, and motion parameter synchronization in R&D mode reduce information loss between design and development.

Engineering Features:

• Full stack vector engine: Supports W3C standard vector editing (3D rotation, bevel, enhanced shape), and the dynamic layout system uses Grid + percentage positioning to ensure multi-terminal adaptation accuracy.
• Underlying performance optimization: v1.16 greatly reduces memory usage, v1.28 introduces canvas paging loading to support large document editing, v1.35 optimizes image decoding memory usage.
• MCP Toolset: 18 standardized tools cover query, editing, design system, resource export and other scenarios. AI Agent can perform 6 operation types such as insert, copy, update, move, delete and image generation through batch_edit. A single call supports up to 25 operations. capture_screenshot and capture_layout provide visual feedback for AI, forming a concept of "observation, understanding, operation, and verification".

Applicable scenarios: Production, design and research teams that need rapid prototype verification, multi-person collaborative design, design system management, and design draft conversion to code can best benefit from Ardot's AI full-process embedding.

How to use Ardot

Multi-terminal access

Access method	Applicable scenarios	Obtain method
Web side	Quick trial, no installation required	Browser visit https://ardot.tencent.com/
macOS desktop client	Available offline, complete font management	Official website download (Apple Silicon / Intel)
Windows desktop client	Full platform coverage	In internal testing, coming soon
iOS / Android	Mobile viewing and comments	App Store / App Store
MCP Server	AI Agent programming control design file	Enable MCP configuration in macOS desktop

Beginner’s Guide

1. Registration and Login: Supports WeChat scan code, mobile phone number/email QQ to log in to iOA enterprise login.
2. Create workbench: After entering the workbench, you can create a new file in the "Draft" personal space or the "Project" team space. Ardot supports direct import of Figma (.fig) files for zero-cost migration.
3. AI-assisted creation: Press Cmd/Ctrl + \ in the editor to open the AI ​​assistant, enter a natural language description (such as "Create a registration page for a fitness app"), and AI will generate a first draft based on the team component library.
4. Professional editing adjustment: Use vector tools (R Rectangle O Ellipse T Text P Pen) and automatic layout function to make fine adjustments to the AI ​​generated results.
5. Collaboration and Sharing: Click the share button to set access permissions and invite team members to collaborate in real time. Development students can view annotations and copy code in R&D mode.
6. MCP Automation (optional): Enable MCP Server configuration on the macOS desktop to support AI Agents such as CodeBuddy, Cursor, and Claude Code to directly read and operate design files.

MCP Server typical workflow

Ardot MCP Server can be started with one click through the built-in configuration portal on the desktop, eliminating the need to manually edit the JSON configuration file. Supports mainstream AI agents such as CodeBuddy, WorkBuddy, Cursor, Claude Code, and Codex. Once configured, the AI ​​Agent can interact directly with the design files through 18 tools:

User: Help me generate React component code based on the login page of the current design draft
AI calls fetch_editor_state to get the current state
AI calls capture_screenshot to obtain page screenshots
AI calls batch_read to obtain the layer structure and attributes
AI calls fetch_guidelines to obtain code generation specifications
AI generates complete React component code

Product Pricing for Ardot

The pricing model is subject to the official real-time page. Usually a freemium or subscription system is adopted. Basic functions can be used for free, while advanced functions or high-frequency use require paid subscriptions. It is recommended that users evaluate the optimal solution based on actual usage.

Application scenarios of Ardot

Scenario 1: Full-process collaboration between production, equipment and research (core scenario)

Product managers use AI to quickly generate prototypes, designers refine visual drafts in professional editors, and R&D developers directly obtain code through R&D mode or MCP. The entire process is completed within a single platform of Ardot, reducing the friction of switching between multiple tools such as Figma + design draft hosting + code generation. Benefits: The cycle from UI design to development delivery is shortened from days to hours.

Scenario 2: Mass production of marketing visuals

The operations team uses Ardot's AI text/image generation, batch generation of vector icons, and intelligent frame selection and modification capabilities to quickly produce visual materials such as social media graphics, event posters, and e-commerce banners. Combined with the team component library, the visual style is always unified. Benefit: Design time for a single visual material dropped from 1-2 hours to 10-30 minutes (deduced value).

Scenario 3: Design system management and maintenance

Through Ardot's design variable system (color, spacing, font size, etc. Design Token), component system (main component + instance + variant) and team library publishing mechanism, the design team can establish unified design specifications and reuse them across multiple product lines. Variable linkage global modification, one adjustment can update all references. Benefits: Design system maintenance costs are reduced by more than 60% (deduced value), and visual consistency across product lines is significantly improved.

Scenario 4: AI empowered automated workflow (MCP)

The development team uses Ardot MCP Server to connect the design files with the AI code generation tool. Typical link: AI Agent reads the structure of the design draft, understands the layout and style, automatically generates front-end code, and writes back the design file. Suitable for agile development teams that require "design as code". Benefits: The docking cost from design draft to code restoration is reduced by about 70% (deduced value), reducing the "pixel-level restoration" work of front-end engineers.

Applicable groups of Ardot

• Product Manager: Use AI to quickly prototype, quickly verify multi-style solutions, and reduce communication costs with designers. You need to master basic AI prompt word skills.
• UI/UX Designer: AI assists in drafting and automating repetitive operations, allowing designers to focus more on creativity and experience decision-making. For designers who accept AI as a "design co-pilot" rather than a replacement.
• R&D Engineer: Directly obtain design annotations and codes through MCP or R&D mode, reducing the duplication of "pixel adjustment against the design draft". Need to understand MCP configuration and AI Agent usage.
• Operations and Marketing Staff: Produce social media graphics and event posters in batches without relying on a design team. Suitable for scenes where visual quality requirements are "sufficient and fast".
• Small and medium-sized enterprise design team: Need to manage the design system and collaborate with multiple people, but the budget is limited. Team Edition's limited-time free policy reduces initial costs.

Does not fit boundaries:

• High originality professional design: For scenes that require extremely sophisticated illustrations and brand visual identity system (VIS) design from scratch, the controllability of AI generation is still limited. It is recommended that professional designers use Illustrator or Figma to create manually.
• Serious publication design: Ardot's automatic layout and AI generation may not be able to meet the requirements for extremely high format accuracy, such as book layout and magazine design.
• Large-scale enterprise purchases need to be prudent: The price of the Enterprise version has not been disclosed. Although enterprise-level capabilities such as SSO/SCIM have been listed, the actual deployment maturity needs to be confirmed with Tencent Cloud Commerce. Data localization and compliance need to be evaluated in conjunction with the company's own requirements.

Summary and Outlook of Ardot

Ardot's core competitiveness lies in its positioning as an "AI native design platform" - it does not superimpose AI functions on traditional design tools, but embeds AI into the entire design process from the bottom up. This gives it significant efficiency advantages in rapid prototype generation, design system management, and conversion of design drafts to code. In particular, the introduction of MCP automation provides engineering infrastructure for the "design as code" workflow.

Current Limitations:

• The product is relatively young, and its ecological maturity (plug-in market, third-party integration, community templates) cannot yet match Figma.
• The Windows desktop version is still in internal testing, and full platform coverage has not yet been completed.
• The actual usage cost of the AI ​​Credits measurement system needs to be paid attention to after exceeding the free limit, especially in high-frequency AI operation scenarios.
• Market recognition and third-party ecosystem are still on the rise, and this stage is more suitable for innovative teams willing to embrace new tools.

Procurement and Adoption Risk Assessment:

• Pilot strategy: It is recommended to use the free version of Starter to verify whether Ardot's AI capabilities and collaboration efficiency meet actual needs in a small-scale team. Focus on AI generation quality, MCP stability, and compatibility with existing workflows.
• Extension Conditions: When the team confirms that AI-assisted design can improve efficiency by more than 30% (based on specific projects), and team members adapt to the new tools, they can be upgraded to the Team version for full rollout.
• Key terms that enterprises need to verify before purchasing: (1) Copyright ownership and commercial authorization boundaries of AI-generated content; (2) Credits usage cap and excess fee limit; (3) Data storage region and compliance certification (SOC2/GDPR, etc.); (4) Actual deployment complexity of SSO/SCIM integration; (5) Completeness of separation asset recovery and data export mechanisms. Since the price and terms of the Enterprise version require business confirmation, it is recommended to sign a clear SLA with Tencent Cloud before purchasing.

Ardot represents the evolution direction of design tools from "manual + template" to "AI driven + automation". For industry, design and research teams who are willing to try new workflows, it is one of the most deeply integrated AI choices among domestic design tools.

Related tools: midjourney, stable-diffusion

How to use Ardot

• Web client: You can use it by visiting the official website and registering an account. Most functions do not require installation.
• API access: Provides RESTful API, developers can obtain the API Key and integrate it into their own applications.

Version Info

• Ardot v1.51 ：Automatically fall back to web rendering when the GPU is incompatible, improve the opening speed of large files, add a new point-added package sales board, optimize AI assistant lag and connection stability, and fix a number of MCP and editing defects. （2026-07-01）
• Ardot v1.50 ：Optimize the CDN storage strategy of PAGX for file preview, optimize the stability of AI assistant MCP connection, and fix multiple rendering and editing defects. （2026-06-30）
• Ardot v1.49 ：The mini program has added a new homepage and message collaboration entrance. MCP supports image upload and download, supports dragging image files to the canvas, and optimizes desktop performance and macOS compatibility. （2026-06-26）
• Ardot v1.48.194 ：The dark mode is fully launched, supports the combined login of mobile phone number and WeChat account, the variable table supports search and filtering, and optimizes the copy and paste performance and the stability of the AI ​​assistant. （2026-06-22）
• Ardot v1.19.91 (AI assistant online) ：The AI ​​assistant is newly launched, supporting batch generation of SVG icons, Vincent graphics, and UI generation based on local component libraries. Internal beta users can receive 1,500 Credits for free. （2026-04-21）
• Ardot is online ：Tencent launches Ardot, an AI intelligent design tool that provides AI-assisted visual creation and vector editing capabilities; there is no official precise date yet. （~2025）