Designing with Leo, BioRender’s AI Agent

Leo by BioRender is our AI agent that is built on BioRender’s figure-building expertise. Leo plans the layout with you, builds from your approved sketch, and keeps every element editable. 

This article explains how to start a figure with Leo, work through the planning and sketch stages, and convert the finished figure into fully editable elements.

Overview

Every Leo chat will guide you through the following phases, with simple opportunities for modification every step of the way:

Initial description → Planning Questions → Sketches → Color draft → Fully editable figure

How to get started with Leo

1. From your gallery, click “Create New”.
2. Select “Illustration”.
3. In your canvas, click “Leo” on the left navigation bar.

4. You can get started in one of two ways:
   1. Use a guided prompt: Leo walks you through it. Best if you're not sure where to begin.
   2. Write your own prompt: Type your full description into the chat. Best if you already know what you want.

Note: Each slide is in your canvas one session with Leo.

Step 1: Initial Description

Option A: Use a guided prompt

Pick one of these under Need help getting started?:

Start from an idea: For when you have a concept but no figure or file.

1. Select “Start from an idea”.
2. Pick a figure type: Protocol/workflow, Mechanism of action, Visual overview, Study design or Other.
3. Edit the sample prompt Leo adds, then click “Send” (↑).

Remake a figure: Recreate an existing figure in BioRender style.

1. Select “Remake a figure”.
2. Click “Add figure” and upload a JPEG or PNG.
3. Finish Leo's prompt with any additional changes, then click “Send” (↑).

Start from a reference file: Build a figure from a paper, deck or image.

1. Select “Start from a reference file”.
2. Pick what to make: Summarize visually, Figure for a slide or Image as reference.
3. Click “Add paper, deck or image” and upload your file (PDF, PPTX, DOCX, JPEG, PNG, CSV or XLSX).
4. Add any details you'd like (optional), then click “Send” (↑).

Option B: Write your own prompt

1. Type what your figure should show and any details Leo must include.
2. To give Leo a file to work from, attach it with the paperclip icon (optional).
3. Click “Send” (↑).

Example: "Make me an MOA figure showing how the brain can adjust myelin on neurons during learning. Show how changes in myelin affect signal timing so that signals from different neurons arrive at their target at the same time."

Tips

• Start simple. Focus on the main idea. You can add detail at every stage.
• Don't worry about layout. Leo asks about it in the next step.
• Change your mind? Click the X next to an option to go back.
• To revisit an earlier step, click its blue box in the chat.

Step 2: Planning Questions

Before it draws, Leo asks a few questions about your figure, such as how to structure it and what to highlight. The questions open in a Figure brief panel.

• Pick an answer: Answers marked Leo's take are Leo's suggestions. Use your own knowledge of the science to choose.
• Add context (optional): Type extra details under any question.
• Skip a question: Leave it blank and Leo will decide.

When you're done, click “Looks right”. To let Leo answer the rest, click “Decide for me”.

Step 3: Sketches

Leo makes four rough sketches based on your answers. Sketches are grayscale and show only the layout, so you can focus on structure before color is added.

Pick the sketch that tells your story best, or click “Decide for me” to let Leo choose.

Optional: Review and mark up your sketch

• See it full size: Click a sketch. Use the arrows to move between sketches, then click “Back” to see all four.
• Draw on it: In Markup mode, cross out what you don't need, draw a missing step, or circle what to emphasize.
• Add comments: Switch to “Comment mode”  to type specific changes.

When you're ready, click “Select and submit markup”. Leo uses your sketch and marks to create the color draft, and shows its progress and reasoning at every step.

Leo keeps you updated on progress and reasoning every step of the way.

Step 4: Color draft

Once the sketch is approved, Leo creates the figure in a color draft. Leo applies consistent icon style, scale, spacing, label hierarchy, and arrow routing based on the approved sketch.

From the color draft, you can continue to make mark-ups and edits as needed.

Once you’ve applied all edits, you can choose to add the figure to your canvas.

Step 5: Add your editable figure

When you're happy with your figure, choose one and then click “Add to canvas”:

• Add as flat image: Adds the figure as a single image. You can make it editable later.
• Make editable: Turns every label, arrow, icon and shape into its own object. You can move items and add icons, shapes and lines.

Once your figure is editable, you can fix a label, swap an icon or move an arrow on the canvas just as you would in BioRender. You can also ask Leo to make the change. Editing one item doesn't change the rest of the figure.

Step 6: Share your feedback

When your figure is editable, Leo asks for feedback in the chat. Your answers help us improve Leo.

1. Rate how useful the figure is (1–5 stars).
2. Rate your overall experience with Leo (1–5 stars).
3. Tell us what worked and what didn't (optional). For example, scientific accuracy, layout, how easy it was to edit, or the chat itself.
4. Click “Share feedback’.

To skip it, click “Not now”.

Leo features

Leo includes several features to make it easier to manage your figure-building conversations:

• Attach a file: Select the paperclip icon to upload a file for Leo to reference when creating or rebuilding your figure.
• Dictation: Select the microphone icon to describe your figure or provide instructions using your voice.
• Start a new chat: Select the chat icon in the upper-right corner, then select “New chat” to begin a new conversation with Leo.
• Adjust text size: Select the three-dot menu to change the chat text size. Choose from Extra small, Small, Medium, or Large.

Use these options at any point while working with Leo to provide additional context, start a new figure, or adjust your chat experience.

Best Practices & Tips

• To revisit an earlier step, click its blue box in the chat.
• Start in BioRender! Avoid copy/pasting AI-generated very long, detailed prompts. Leo will guide you to focus on the purpose of your figure, supporting you through your creation process.
• You can review and mark up the figure at every stage - the plan, the sketch, and the finished figure. Leo updates only what you flagged.
• Leo asks about intended use and audience early on; the more specific your answers, the closer the first sketch will be to what you need.
• Start from an existing figure. Upload a reference figure or use one already on your canvas and ask Leo to rebuild it as-is or with specific changes. You don’t have to start from scratch.

Troubleshooting

If the generated figure wasn’t what you expected, try the following:

• Make the figure editable to fine-tune elements like labels and arrows
• Use more specific or detailed language in your prompt
• Break complex workflows into multiple figures
• Provide an example of an intended layout

If you reach the Leo conversation limit but still want to continue editing, try the following:

• Copy and paste the generated image into a new slide on your canvas, then start a new chat with Leo to refine further
• Export the generated image, then start a new chat with Leo with the image attached
• Use BioRender’s Custom Edit option in the AI toolbar

If you don’t see “Leo” in your canvas

• Account access: Leo is not available on BioRender Basic (free) accounts. If you’re on a free account, the Leo icon will not appear in your canvas.
• Browser settings: Your browser may be blocking services required for Leo. If you use an ad blocker or browser privacy extension, disable it for app.biorender.com, then refresh your browser.

Frequently Asked Questions

What is Leo? 

Leo is the first AI agent built for scientific figures. Built on BioRender's figure-making expertise, Leo plans the figure with you, builds it from your feedback, and keeps every label, arrow, and icon editable.

How is Leo different from tools like ChatGPT or Gemini? 

General AI image tools turn a prompt into a flat image. Leo is built on BioRender's figure-making expertise. It asks about your figure first, shows you a sketch to mark up, and then builds the figure in BioRender. Every label, arrow, and icon is a separate object you can edit.

Do I need to know how to lay out a figure before using Leo? 

No. You describe what the figure needs to say, and Leo asks the questions that determine the layout - intended audience, use case and level of detail.

Can I edit the figure after Leo builds it? 

Yes. Once converted, every label, arrow, icon, and shape is a real, selectable object on the canvas. You can edit any element yourself or ask Leo to change it.

What is the limit on conversations with Leo? 

Leo can support a maximum of 10 messages per conversation, but new chats can be started to pick up where you left off.

Does editing one part of the figure affect the rest? 

No. You can change one element (i.e. a label, an icon, or an arrow) without regenerating the rest of the figure.

Is my data used to train AI models?

No. Your data is never used to train AI models without consent.

How much does Leo cost to use?

To use Leo, users must have an AI Add-on plan. Certain Leo actions cost AI Credits:

• Generating a draft figure
• Generating sketches beyond the first round of sketches

Other Leo actions are free:

• Asking Leo questions (initial chats with Leo)

How can I avoid copyright issues when using BioRender AI?

Start with BioRender's own icons and templates, or with content you already own. This is the safest way to use AI features, because BioRender's icon library is vetted for use in your figures.