Anima Agent and Figma: AI Redefines Interface Design

The Future of Interface Design: Agentic Design

The year 2026 is shaping up to be a pivotal time for product design, marked by the integration of artificial intelligence (AI) into the design process. While in the past, AI primarily served as an assistant, it is now an active partner in creating user interface designs. This evolution has transformed how designers approach their work, shifting from simple assistance to close collaboration with AI.

Two distinct approaches are emerging in the creation of user interface designs. The first involves allowing AI to generate a complete design, from the initial concept to the coded prototype, a method known as AI Autonomous Design. The second, more traditional approach involves using AI tools to create a basic design that designers can then refine in software like Figma.

Anima Agent: An Innovative Tool for AI-Assisted Design

Among the promising tools in the realm of AI-assisted design, Anima Agent stands out for its approach. This plugin for Figma enables the creation of sophisticated user interface designs by following the second approach mentioned earlier. To use Anima Agent, simply select the Actions option in Figma and search for Anima Agent. Once launched, the tool prompts the user to submit a prompt to generate a user interface design.

Anima Agent defaults to using the Claude Sonnet 4.6 AI model, but it is possible to choose other models, such as OpenAI GPT 5.3 Codex. According to user experiences, Claude Sonnet 4.6 works effectively in 95% of cases.

Note on Claude 4.6 Opus

Opus is a more advanced AI model developed by Anthropic. Although it consumes more tokens on average, it offers output quality similar to that of Sonnet 4.6. It is recommended to choose this model only if you have a specific reason to do so.

Scenario 1: Creating a New Design

The process of creating a new design with Anima Agent is relatively straightforward. The user must describe the desired project in the form of a prompt and submit it to the Anima AI agent. For example, to design a mobile application for daily health tracking, the prompt could include specific UX and visual requirements, such as data clarity and a modern design.

🎯 Task

The task is to design a mobile application that helps users track and understand their daily health indicators.

👤 Target Users

Target users include individuals looking to improve their health habits, those tracking their sleep, steps, hydration, and overall well-being, as well as non-technical users who appreciate simplicity and clarity.

📱 Main Screens to Design

The main screens to design include:

1. Onboarding (goal selection and customization)
2. Main dashboard (daily overview)
3. Activity tracking (steps, movements, calories)
4. Sleep tracking
5. Hydration tracking
6. Insights & trends (weekly/monthly health data)
7. Profile / Settings

🧠 UX Requirements

The UX requirements include prioritizing clarity and low cognitive load, displaying the most important data first (progress rings, summary cards), using progressive disclosure for detailed data, including micro-interactions (animations, transitions), and providing motivating elements (series, goals, gentle encouragement).

🎨 Visual Design Requirements

The visual design requirements include a clean and modern Apple-quality design, a soft color palette (health-oriented: greens, blues, neutrals), high contrast for accessibility, the use of a card-based layout with subtle shadows, rounded corners, generous spacing, and a clear and readable typographic hierarchy.

📊 Components to Include

The components to include are progress rings (daily goals), line graphs (health trends), cards (metrics like steps, sleep, hydration), toggle switches and input controls, and a bottom navigation bar.

⚙️ Design System Guidelines

The design system guidelines include a consistent spacing system (8pt grid), reusable components, a defined typographic scale, and accessible color contrast.

🚀 Expected Outcome

The expected outcome is a detailed description of each screen, including layout structure (top → bottom), interaction behavior, visual style direction (colors, typography, spacing), and suggestions for component states (active, empty, loading).

⚠️ Constraints

The constraints include avoiding generic or template-like user interfaces, steering clear of overly complex layouts, focusing on true usability (not just visuals), and designing for an iOS-first experience.

When using this tool, it is common to request the creation of 5 to 7 main pages/screens representing the site or application. The goal is for all screens to have a consistent look and feel, making them easy to use for demonstration purposes.

Anima Agent keeps the user informed of its progress throughout the process. For example, the tool may indicate that it is currently working on generating images for the user interface layout.

After a short delay, typically a few minutes, the user receives a series of screens. The result is often satisfactory for a first attempt.

A notable advantage is that the design has built-in automatic layout from the start, making modifications easier.

However, it is crucial to mention that the design generated by Anima may have visual flaws. For instance, in one case, the bottom tab bar was positioned in the middle of the screen for no apparent reason.

The good news is that these issues can be easily corrected, as the design is available in Figma. A final check is still necessary before passing the design to the development team.

Scenario 2: Creating Variants of an Existing Design

To create a variant of an existing design, the user must select the frames containing the current design and submit a prompt to Anima specifying the desired changes.

For example, the prompt could be: Create two other variants of this screen. Use different types of data visualization on this screen.

Anima will generate new variants and place them side by side with the existing designs.

Quick Tip

It is recommended to focus on a specific aspect of the design when creating variants. For example, the user might ask Anima to create different data visualizations.

Scenario 3: Creating Design Using Your Own Design System

Creating a design from scratch is effective for the exploration phase of the design process. However, when the organization has a design system in place, it is often preferable to use it to develop the design. Fortunately, Anima can assist in this process. The user simply needs to choose the components to use and ask Anima to create a specific design by integrating them.

In one example, the user might select 4 components (Title, Input Field, Checkbox with Label, and Call-to-Action Button) and ask Anima to create a login page.

The prompt could be: Build a login form with two fields (username and password), a checkbox ("Stay logged in"), and a call-to-action "Log in."

Anima then generates a visually appealing login form, reusing the 4 components in the final design.

However, one issue must be mentioned. Although the components look identical, they are actually detached elements and not real Figma components. Thus, when clicking on the input field, the user discovers that it is essentially a frame with the visual properties of the original component.

Conclusion

Anima Agent represents a significant advancement in user interface design, allowing for a swift transition from idea to implementation while providing the opportunity to refine generated designs. The tool does not take control of the creative process but enriches the designer's experience by facilitating the integration of AI into the workflow.