The Complete Carosello Walkthrough: Create Professional LinkedIn Carousels in 5 Minutes

Creating LinkedIn carousels that capture attention and drive engagement used to require hours of design work, expensive tools, and professional skills. With Carosello, you can transform your ideas into stunning, professional carousel posts in just 5 minutes — no design experience required.

This comprehensive walkthrough takes you through every feature of Carosello, from your first login to exporting a polished carousel ready for LinkedIn. We'll cover content generation, slide editing, AI image creation with dual-reference controls, prompt reverse-engineering with Image to Prompt, precision retouching, brand customization, and everything you need to create carousels that get results.

Why LinkedIn Carousels Outperform Other Content Formats

Before diving into the walkthrough, here's why carousels deserve your attention:

The engagement advantage is real:

• Carousels generate 3-5x more engagement than standard text posts
• Users who swipe invest more attention, creating deeper connections
• The storytelling structure (hook → value → CTA) keeps readers engaged
• LinkedIn's algorithm rewards content that keeps users on the platform longer

The time investment problem:

Traditional carousel creation requires:

• 15-30 minutes with design tools like Canva (if you have design skills)
• 30-45 minutes with professional tools like Figma
• 1-2 hours if you're starting from scratch without templates

Carosello solves this by combining AI content generation, professional image creation, and intuitive editing — all in one platform built specifically for LinkedIn carousels.

---

Getting Started: The Carosello Dashboard

When you first access Carosello at carosello.app/app, you'll see the main dashboard — your command center for carousel creation.

Carosello landing page with prominent Start Creating call-to-action button

What you see on the dashboard:

The dashboard is intentionally minimal — your focus should be on creating content, not navigating complex interfaces.

Action: Click the "Start Creating" button to open the new post configuration modal.

---

Step 1: Creating a New LinkedIn Post

The "New LinkedIn Post" modal is where your carousel journey begins. This is where you tell Carosello what you want to create and provide any context that helps the AI understand your content goals.

The Empty Form

New LinkedIn post modal showing empty configuration form

The modal presents a clean, focused interface with essential configuration options:

Required fields:

• What should I create? — Describe your carousel topic, goal, or the story you want to tell
• Number of Slides — Select 1-6 slides for your carousel

Optional enhancements:

• Language — Choose the output language for your content
• Reference Documents — Upload PDFs, TXT, or MD files for additional context

Filling Out the Form

Let's walk through creating a carousel about scaling a marketing team — a topic that resonates with founders, marketing leaders, and professionals navigating growth.

New post form filled with prompt about scaling marketing teams and 5 slides selected

Best practices for your prompt:

1. Be specific about the outcome — Instead of "marketing tips," try "3 frameworks for scaling a marketing team from 5 to 50 people"
2. Include your angle — What makes your perspective unique? What experience informs this content?
3. Define your audience — "For Series A founders" or "For marketing leaders at scale-ups"
4. Specify the value — What will readers learn, gain, or be able to do after reading?

Example prompt that works well:

"Create a carousel about scaling marketing teams from 5 to 50 people. Include the three phases of team growth, common mistakes to avoid, and actionable frameworks. Target audience: startup founders and VPs of Marketing at Series A-C companies. Tone: practical, experienced, anti-guru."

Adding Context with Reference Documents

For content that requires specific data, frameworks, or detailed information, you can upload reference documents. This is particularly powerful when you have existing content you want to transform into carousel format.

Context file uploaded showing Carlos project context document

Supported file formats:

• PDF — Research papers, reports, case studies
• TXT — Plain text notes, transcripts, outlines
• MD — Markdown files, structured content, documentation

When to use reference documents:

• Transforming a blog post into a carousel
• Creating content from research or data
• Ensuring accuracy when citing specific frameworks
• Maintaining consistency with existing content

The AI analyzes your uploaded content and incorporates relevant information into the generated carousel, ensuring accuracy and depth.

Action: Click "Continue" to start the AI generation process.

---

Step 2: Watching AI Generate Your Content

After clicking Continue, Carosello's AI goes to work. The generation process typically takes 15-30 seconds, depending on the complexity of your request and the number of slides.

LinkedIn post preview showing content generation in progress

What happens during generation:

1. Content Analysis — AI processes your prompt and any uploaded documents
2. Structure Planning — Determines the optimal narrative flow for your slides
3. Hook Creation — Generates an attention-grabbing opening slide
4. Value Delivery — Creates content slides that deliver on your promise
5. CTA Crafting — Develops a compelling call-to-action for the final slide

The progress indicator keeps you informed throughout the process. You'll see each slide being generated in sequence.

---

Step 3: Reviewing Your Generated Post

Once generation completes, you'll see your full carousel with AI-generated content, including a narrative analysis that explains the structure and strategy behind your post.

Generated LinkedIn post with narrative analysis showing content structure and strategy

The generated output includes:

The Narrative Analysis: Your Content Strategy Partner

The narrative analysis is your secret weapon. It reveals the AI's thinking process and gives you a detailed breakdown of the content strategy.

Narrative structure analysis panel showing hook analysis, content structure, and rationale

Here's what the analysis surfaces:

• Hook Analysis — Why the opening line works and what makes it compelling
• Structure Type — Whether the content follows a discovery-led, problem-solution, or framework approach
• Rationale — How the structure maximizes impact within your chosen slide count
• Slide Prompts — Direct links to the image generation prompts for each slide

Take time to review the analysis. It often surfaces insights that help you refine not just this carousel, but your content strategy overall.

---

Step 4: The Slide Content Editor

The Slide Editor is where you refine each slide's content before generating images. This is Carosello's most powerful editing interface, designed to give you complete control while preserving AI-generated insights.

Understanding the Editor Layout

Slide 1 content editor showing AI suggestion panel and content guidance area

Editor components explained:

Progress Bar (Top)

• Shows your current position: "Slide 1/5"
• Click any position to jump directly to that slide

Slide Type Badge (Top-Right)

• Indicates the slide's role: "Opening Hook," "Key Point," "Case Study," "CTA"
• Helps you understand each slide's purpose in the narrative

AI Suggestion Panel (Upper Card)

• The original AI-generated content
• One-click copy button to grab text
• Reference point for comparison

Your Slide Content (Lower Card)

• Editable version of the content
• Undo/redo controls for change management
• This is what gets used for image generation

Navigation Controls (Bottom)

• Back to Post — Return to full post view
• Refine — Request AI improvements to current slide
• Next — Proceed to the next slide

Editing Best Practices

The slide editor preserves your AI-generated content as a reference while giving you a separate space to customize. This dual-panel approach means you can:

1. Compare versions — See the original alongside your edits
2. Experiment freely — Make changes without losing the original
3. Cherry-pick elements — Copy specific phrases from the AI suggestion
4. Maintain consistency — Keep the AI's tone while adjusting specifics

Pro tip: The "Refine" button sends your current content back to the AI with instructions to improve it. Use this when the direction is right but the execution needs polish.

Version History and Diff Comparison

Made changes you want to compare? Carosello tracks all versions and provides a powerful diff viewer with word-level precision.

Unified diff view showing word-level comparison between generated and edited versions

Diff viewer features:

The unified view shows changes inline with full context, making it easy to spot exactly which words were added, removed, or modified. Switch to split view when you want a side-by-side comparison instead.

This is invaluable when you've made extensive edits and want to ensure you haven't lost important elements, or when collaborating and need to see what changed.

Finalizing Slide Content

Once you're satisfied with a slide, click "Next" to proceed. The editor shows your progress and confirms when each slide is ready.

Slide 1 content finalized and ready to proceed to next slide

Completion indicators:

• Progress bar fills as you complete slides
• Completed slides show a checkmark
• You can return to any slide to make additional edits

Completing All Slides

Continue through each slide until you reach the final CTA slide. This slide is particularly important — it's your opportunity to convert engaged readers into followers, commenters, or customers.

Slide 5 showing final CTA content ready to generate carousel images

Effective CTA strategies:

• Ask a specific question — "Which of these 3 phases is your team in right now?"
• Invite action — "Save this for your next team planning session"
• Create conversation — "Drop a comment if you've made mistake #2"
• Offer value — "Follow for more frameworks from 15+ scale-up experiences"

With all slides finalized, you're ready for the most visually impactful step: AI image generation.

---

Step 5: The Image Laboratory

The image generation interface — the Laboratory — is where Carosello truly sets itself apart. This isn't slapping stock photos onto slides. It's a full creative studio for generating, refining, and perfecting custom visuals that match your content and brand.

The Laboratory Interface

Image generation interface showing dual-reference controls for content and style

The Laboratory gives you two working modes:

• Generate & Edit — Create new images from prompts, refine with quick edits, and retouch existing results
• Image to Prompt — Upload any image and extract a detailed prompt from it

Interface components:

LinkedIn Caption Header

• Shows your post caption with best posting time recommendation
• Quick access to edit caption details or export

Prompt Area

• AI-generated image description based on your slide content
• Fully editable to customize the visual direction
• Sanitize button to clean prompts that trigger safety filters
• Improve button for one-click prompt enhancement

Dual-Reference Controls

• Content Reference — Upload images for subject/content matching
• Style Reference — Upload images for visual style matching
• Human Subject Toggle — Enable face-safe mode for portraits

Generation Settings

Aspect ratio, resolution, format, and guidance scale controls

For LinkedIn carousels, 4:5 aspect ratio works best — it takes up maximum screen real estate in the feed.

Selecting Your AI Model

Carosello supports multiple image generation models. For professional LinkedIn content, Gemini 3 Pro delivers the best results.

Image generation with Gemini 3 Pro model selected for professional output

Model comparison:

Use Flash for rapid experimentation. Switch to Gemini 3 Pro for final carousel images.

Your First Generated Image

Click "GENERATE" to create your first carousel image. The AI interprets your slide content and prompt to create a unique, professional visual.

Slide 1 image generated showing revenue curve split graphic visualization

And here's the generated image up close:

Generated slide image showing team scaling infographic with revenue growth curve

What you see after generation:

• Full image preview on the right panel
• Image editing toolbar for refinements
• Status indicator showing "COMPLETE"
• Credit cost displayed (e.g., "4 credits")

Quick Edit: Refine Your Prompt Without Starting Over

Not happy with the result? You don't need to rewrite the entire prompt. The Quick Edit field lets you describe what to change, and the AI rewrites your prompt accordingly.

Quick Edit applied showing refined prompt with subject positioning instructions

How Quick Edit works:

1. Type a short instruction in the Quick Edit field (e.g., "Add the subject in the right bottom corner looking at the chart")
2. Click "Apply tweak"
3. The AI rewrites your full prompt incorporating the change
4. Hit GENERATE to see the result

This is faster than manual prompt editing and keeps the AI-generated structure intact. Use it for iterative refinements — adjust composition, add elements, change mood — without losing what already works.

Image to Prompt: Reverse-Engineer Any Visual

See an image you love and want to recreate in your brand's style? The Image to Prompt tab does the reverse of generation — it analyzes an uploaded image and produces a detailed text prompt describing it.

Image to Prompt interface showing upload area and prompt extraction workflow

How to use Image to Prompt:

1. Switch to the "Image to Prompt" tab
2. Upload or drag-and-drop any reference image
3. Click "GENERATE PROMPT"
4. The AI produces a detailed description of the image's composition, style, colors, and elements
5. Switch back to "Generate & Edit" and use the extracted prompt as your starting point

When this is useful:

• Recreating a competitor's visual style (ethically) in your brand colors
• Maintaining consistency across carousel slides by extracting the prompt from your best image
• Learning what makes a specific visual work by reading its prompt decomposition
• Starting from a mood board image instead of writing prompts from scratch

---

Step 6: Advanced Image Features

Carosello's advanced features give you creative control that typically requires professional design tools. These features transform AI generation from a black box into a precision instrument.

Dual-Reference Image Generation

The dual-reference system is Carosello's most powerful feature for consistent, branded imagery. You provide both a content reference (what to show) and a style reference (how it should look), and the AI combines them.

Dual-reference control panel showing content ref and style ref images attached with Human Subject enabled

Using Subject Reference for Portraits

When creating carousels that feature a person (like a personal brand carousel), subject reference ensures consistency across all slides.

Slide 2 with subject reference portrait uploaded and face-safe mode enabled

To use subject reference:

1. Expand the "Dual-Reference Controls" section
2. Click "DROP CONTENT IMAGE" area
3. Upload a clear portrait or subject photo
4. Enable "Human Subject" checkbox for face-safe mode
5. The image appears with "CONTENT REF" badge

Face-safe mode ensures:

• Facial features are preserved accurately
• The person remains recognizable across different generated scenes
• Professional portrait quality is maintained

Generated Portrait Result

With subject reference and face-safe mode, the AI generates images that maintain the person's identity while placing them in completely new contexts.

Generated Carlos portrait in futuristic office setting matching content and style

Here's the portrait result up close — the subject's features are preserved while the environment matches the professional growth theme of the carousel:

Close-up of dual-reference generated portrait showing subject consistency

Style Reference for Visual Consistency

Style reference controls the aesthetic direction of your generated images. This is how you ensure visual consistency across all carousel slides.

For a tech-forward, professional aesthetic, you might use a style reference with:

• Deep navy backgrounds
• Cyan/purple accent lighting
• Modern, editorial quality
• Tech-optimistic atmosphere

The AI then applies this visual language to all generated images, creating a cohesive carousel that looks professionally designed.

Generated Images Across Slides

Here's how different slides come together with consistent styling:

Slide 3: Three-Phase Timeline

Slide 3 generated showing three phases timeline with professional icons

The timeline visualization maintains the same color palette and aesthetic while effectively communicating the growth phases.

Slide 4: Digital Twin Concept

Slide 4 generated showing digital twin hologram split visual concept

Even complex conceptual imagery — like a digital twin representation — stays consistent with the established visual identity.

Retouch Mode: Precision Image Editing

Sometimes a generated image is 95% perfect but needs minor adjustments. Carosello's Retouch Mode lets you make targeted changes without regenerating from scratch.

Activating Retouch Mode

Click the "Enter retouch mode" button in the image toolbar. The interface switches to editing mode with a crosshair cursor over the image.

Retouch mode toolbar showing instruction input field and action controls

How retouch works:

1. Click "Enter retouch mode" in the image toolbar
2. Click on the area of the image you want to edit — this places a marker
3. Type your edit instruction in the text field (e.g., "remove the text overlay and clean up the background")
4. Click "Apply retouch"
5. The AI processes your request while preserving the rest of the image

Common retouch requests:

• Remove unwanted text or watermarks
• Clean up background elements
• Adjust lighting or color intensity
• Remove distracting objects
• Fix composition issues in specific areas

Retouch Result

Image edit interface showing prompt to remove text overlay from generated image

The unwanted element is seamlessly removed. The background fills naturally, maintaining visual quality:

Clean edited image after retouch with text overlay removed

Retouch credits: Image edits cost approximately 3 credits — much less than full regeneration. Use retouch liberally for quick fixes instead of burning credits on full re-generations.

---

Step 7: Brand Customization in Settings

Before generating your final carousel, configure your brand settings. This ensures visual consistency across all your content and saves time on every future carousel.

Accessing Settings

Navigate to Settings from the main dashboard to configure your brand identity.

Brand Colors

Settings showing full brand color palette overview with multiple color options

Available color settings:

These colors are automatically injected as hex values into your image generation prompts. Change them here, and every future image generation uses your updated palette.

Using the Color Picker

Click any color swatch to open the color picker with gradient selection and hex input.

Color picker open showing gradient selector for precise color selection

The color picker supports:

• Visual gradient selection
• Hex code input for exact matching
• Hue slider for broad color changes
• Quick-pick swatches from your existing brand palette

Making Color Changes

When you change a color, the interface shows unsaved changes with clear save/discard options.

Primary color changed to orange showing unsaved state

Best practice: Test color changes on a single slide before applying to an entire carousel. You can always revert.

Saving Your Brand Colors

Brand colors updated with save and discard buttons visible

Click "Save to Account" to apply changes or "Discard Changes" to revert. These settings persist across all future carousel creations.

Prompt Templates

Prompt templates automatically inject visual identity rules into every image generation prompt. Instead of manually typing "use dark navy background with cyan accents" on every slide, set it once here.

Settings showing prompt templates with visual identity rules configured

Prompt templates can include:

• Visual identity specifications (colors, lighting, textures)
• Style guidelines (photography, illustration, minimalist)
• Brand-specific instructions (always include, never include)
• Safety context and term substitutions (e.g., replace "guru" with "practitioner")

This ensures consistency without manually adding instructions to every prompt.

Brand Profile

Configure your professional profile to give the AI context about who you are and what you stand for. This shapes the tone, perspective, and voice of all generated content.

Brand profile settings showing Carlos Rossetti context and professional background

Brand profile includes:

• Name and title
• Company/role information
• Professional background and expertise
• Content mission and themes
• Positioning statement
• Term substitutions and forbidden phrases for brand safety

API Keys (BYOK)

For power users, Carosello supports Bring Your Own Key (BYOK) for unlimited generation with your own Gemini API key.

API keys settings showing Gemini BYOK configuration with security note

BYOK benefits:

• Unlimited generations (you pay Google directly at their rates)
• Complete cost control and transparency
• No credit purchases required
• Full feature access

Security note: API keys are encrypted and stored securely. They're never visible after initial entry and can be rotated anytime.

Credits and Usage

Monitor your credit balance, purchase credit packs, and review your full usage history in the Credits section.

Credits settings showing balance, pricing tiers, and usage history

Here's the full credits interface showing pricing tiers and a detailed usage log:

Detailed credits view showing balance, pricing packs from Starter to Agency, and recent API usage history

Credit system overview:

Credits never expire. All purchases are one-time payments — no subscription required.

---

Step 8: LinkedIn Caption Management

Your carousel images are only half the story. The LinkedIn caption that accompanies your carousel is equally important for engagement and discoverability.

The Caption Editor

LinkedIn caption editor showing hook, body, CTA sections and hashtag management

Caption structure:

The hook is critical. LinkedIn shows only the first 2 lines before users must click "see more." Your hook must:

• Stop the scroll
• Create curiosity
• Promise value
• Connect to the carousel content

Caption Editing Features

• AI-generated suggestions — Based on your carousel content
• Structured sections — Separate fields for hook, body, CTA, and hashtags
• Best posting time — Recommended posting window displayed in the header
• Copy to clipboard — One-click transfer to LinkedIn or scheduling tools
• Full editor mode — Expanded view for detailed editing
• Engagement tip — AI-generated suggestion for driving comments

---

Complete Workflow Summary

You've now seen every step of the Carosello workflow. Here's the complete process at a glance:

The 5-Minute Carousel Creation Process

Total: 5 minutes from idea to export-ready carousel.

Keyboard Shortcuts

Speed up your workflow with these shortcuts:

---

Tips for Maximum Engagement

Based on data from thousands of LinkedIn carousels, here are proven strategies for better results:

Content Structure

1. Hook on Slide 1 — Question, bold claim, or surprising stat
2. Value on Slides 2-4 — One key point per slide, concrete examples
3. Social Proof on Slide 5 — Case study, result, or testimonial
4. CTA on Final Slide — Specific, easy action to take

Visual Best Practices

• Consistent color palette — Use your brand colors across all slides
• Readable text — If adding text to images, ensure high contrast
• Mobile-first — 70%+ of LinkedIn users browse on mobile
• 4:5 aspect ratio — Takes up maximum screen real estate in the feed
• Professional quality — Gemini 3 Pro for final images, Flash for drafts

Caption Strategy

• Front-load value — Most important info in first 2 lines
• Personal angle — Share your experience or perspective
• Engagement hook — End with a question or invitation to comment
• Hashtag strategy — 3-5 relevant hashtags, mix of popular and niche

Posting Optimization

• Best times — Tuesday-Thursday, 7-9 AM, 12-1 PM, or 5-6 PM
• Immediate engagement — Respond to comments in first 2 hours
• Cross-promotion — Share to LinkedIn story for additional visibility
• Series potential — Create follow-up carousels for engaged topics

---

Troubleshooting Common Issues

Generation Fails

Check first:

• Credit balance (sufficient for the operation?)
• Internet connection (stable?)
• Prompt complexity (try simplifying)

Quick fixes:

• Refresh the page and retry
• Use a shorter, more focused prompt
• Check status page for any service issues

Image Quality Issues

For better results:

• Select 2K resolution for final images
• Adjust Guidance Scale (lower = more creative, higher = more literal)
• Provide clearer reference images
• Be more specific in your prompt

Safety Filter Blocks

If generation fails with a safety filter block:

• Click the "Sanitize" button to automatically clean triggering words from your prompt
• Check Settings > Brand Profile for term substitutions (e.g., "guru" → "practitioner")
• Try switching to Gemini 3 Pro, which handles complex prompts more gracefully
• Retry after sanitizing — most blocks resolve immediately

Version Diff Not Showing

Requirements:

• Multiple versions must exist (make edits first)
• Try different comparison modes (Words/Chars)
• Switch between Unified and Split views
• Refresh the page if display issues persist

---

What's Next?

You now have everything you need to create professional LinkedIn carousels with Carosello. Here's how to continue:

Immediate Actions

1. Create your first carousel — Apply what you've learned
2. Configure brand settings — Set up colors and templates
3. Test different approaches — Experiment with prompts and references

Ongoing Optimization

• Analyze performance — Track which carousels drive most engagement
• Refine your style — Develop signature visual elements using style references
• Build a prompt library — Save successful prompts and use Image to Prompt to extract patterns from high-performing visuals

Advanced Features to Explore

• BYOK setup — For unlimited generation with your API key
• Prompt templates — Automate brand consistency across every image
• Retouch workflows — Generate fast with Flash, then retouch with precision
• Image to Prompt chains — Extract a prompt from your best image, tweak it, and generate variations

---

Start Creating Today

You've seen how Carosello transforms the carousel creation process from hours of design work into a 5-minute workflow. Professional content, stunning visuals, and complete creative control — all in one platform built specifically for LinkedIn success.

Ready to create your first carousel?

Every feature covered in this walkthrough is available right now. Upload your content, let AI generate your post and images, and export your complete LinkedIn Content Package in minutes instead of hours.

Start Creating Free →

No credit card required. Full feature access. Create your first professional LinkedIn carousel today.

---

This walkthrough covers Carosello version 1.0. Features are continuously updated to improve your carousel creation experience.