Seeing the future of design: Lovart's world's first design agent experience

We have been internally testing an amazing Agent product recently, and today it was finally officially announced.

It's called Lovart, and it's the world's first professional design agent.

After the launch of the GPT 4o raw image function, the field of raw images has entered a new era.

The significant progress of the text graph model also provides new possibilities for the agent-based design tools.

So what exactly is a design agent?

Let's watch an official video to get a feel for it:

1. Full-link design and execution, done in one sentence

Previous text-generating image tools provided the task of “generating images”.

The design agent is like a "design executive", covering the entire visual process from creative analysis to professional delivery.

From intention breakdown → task chain → final product, it’s all done in one sentence.

Hundreds of steps can be executed at a time, and up to 40 finished design drawings can be produced at a time.

For example, I can ask it to make a set of silver gradient emoticons based on photos of my cat. It will design the expressions on its own, take pictures one by one, and edit them all, making black-background gradient versions of all the emoticons with just one click.

Prompt:

Use the cat in this picture as a template to design a set of 12 silver gradient emoji packs. Style 3D cartoon, to express the most popular and trendy emoji this year.

For another example, create a 6-frame storyboard for a 30-second video ad promoting a smart home speaker.

It will first write the story according to the needs, and then start character design.

After completion, the design of key scenes will be carried out.

Finally, fill the scenes into the storyboard, and add information such as lighting and camera movement to each picture.

Prompt:

Create a 6-frame storyboard for a 30-second video ad promoting a smart home speaker.

Scene progression:

Modern living room, family interaction

Voice command to the speaker

Music starts playing

Speaker answers a question

Lighting adjusts automatically

Close-up of product and brand tagline

Style: clean, semi-realistic illustration, soft light.

2. Image, video and music, free scheduling

Only with a sufficient toolbox can the Agent complete the task more comprehensively.

Lovart integrates a series of cutting-edge AI models, such as GPT image-1, Flux pro, OpenAI-o3, Gemini Imagen 3, Kling AI, Tripo AI, Suno AI, etc.

Designers can complete design, editing and generation in a unified interface, no longer having to switch back and forth between multiple software and models.

I asked it to design a new lifestyle tech brand "VIVIDPIX" and create a cross-media launch campaign.

It not only designed the logo, posters, and peripherals, but also used Keling to produce a 3D version of the logo animation and used Suno to produce the background music.

Since all models have the same Context, the coordination is seamless.

Prompt:

Prompt:

Create a cross-media launch campaign for a new lifestyle tech brand called “VIVIDPIX”.

Visuals: Design two main brand posters in different styles (futuristic minimalism and playful retro), including editable text layers and logo integration.

Video/3D: Generate a dynamic 10-second video loop of the brand logo morphing in 3D space, with smooth, modern animation, suitable for website and social sharing. Leverage Kling or equivalent video/animation AI tools.

Music: Compose an original 15-second catchy theme music that matches the brand's energetic and optimistic tone. Use Suno or other advanced AI music models.

Packaging: Output a box and product mockup, harmonizing visuals across all touchpoints.

Canvas: Bring all assets together onto a single editable Lovart canvas, allowing for integrated fine-tuning and end-to-end brand asset delivery.

All elements—visuals, video, music, packaging—should have a cohesive color palette and brand personality, and be ready to use for a multi-platform product launch.

3. Generate results automatically in layers and modify them as you like

In the generation of posters, the biggest pain point is the post-editing of texts, and Lovart solves this pain point very well.

For example, here I am designing a promotional poster for a new plant protein drink.

A minute later, it generated the standard poster on the left, with a separate version of the image and text on the right.

Now I can freely modify the content, size and color of the text, which is really convenient.

Prompt:

Design a promotional poster for a new plant-based protein drink.

Style: vibrant and energetic, modern editorial layout.

Color palette: green, cream, and orange.

Text: “Fuel Your Day, Naturally”.

Aspect ratio: 3:4, for Instagram feed.

Include editable text layers.

Tips for use:

I also discovered some tips while using Lovart, which I would like to share with you:

1. 1.  Lovart has many built-in image, video, and music models, but not every model supports Chinese well enough, so using English prompts will provide the best experience.
2. 2.  The starting point of design is demand. Only when the demand is well described can the Agent be designed well. Therefore, it is recommended to discuss the demand with AI before issuing the task and let AI output the final prompt word. The prompt word can contain structured information such as style, color, text, and proportion. The structured information itself is a decomposition of complex requirements. The prompt words written in this way will have the best effect when Lovart executes.
3. 3.  If you need a separate text layer, just add this sentence to the command: Include editable text layers.

Conclusion

The emergence of Lovart represents that the role of AI in the field of design is changing from a simple "image generator" to a deeper "design executive".

What it tries to solve is not just the problem of "drawing a picture" but the problem of "completing a design task".

Someone once asked me if the advent of 4o Raw would kill the image workflow.

As AI agents begin to understand and execute more complex design processes,

We and AI will also usher in a new way of collaboration.