How to Automate Instagram Carousels With AI (Viral Templates)
How I automate Instagram carousels with AI using Make.com or n8n, Airtable, and Blotato's viral template library. Three industry builds, end to end.
If you want to automate Instagram carousels without designing every slide by hand in Canva, this is the exact workflow I run, with three industry-specific examples baked in. Most “AI carousel” guides stop at one ChatGPT prompt and a free template pack. This one shows you how to wire OpenAI, a viral template library, Airtable, and a publisher into a single automation that spits out finished carousels for real estate, life coaching, or restaurants on autopilot.
Below are two builds, one in Make.com and one in n8n, with three industry branches each. Pick the tool you already use, then copy the branch that matches your niche.
How to Automate Instagram Carousels (Video Guide)
If you’d rather watch the full walkthrough, this is the video version with Kevin Farrugia, who built the Make.com and n8n blueprints. The written guide below covers the same three industry branches with extra detail on the OpenAI prompt step, the wait-loop pattern in n8n, and the gotchas that break carousel automations.
Why Most Carousel “Automation” Guides Don’t Actually Automate
Search “automate Instagram carousels” and you get two flavors of advice. One is “open Canva, duplicate a template, type your slides.” That’s manual work with extra steps. The other is “schedule a finished carousel in Later or Buffer.” That’s scheduling, not automation. Neither one solves the part that actually takes time: writing the slide copy and designing each slide.
Real automation means the workflow writes the copy, generates the visuals, and queues the post. You feed it one trigger (a topic, a row in a spreadsheet, a scheduled run) and a finished carousel comes out the other end. That’s what this build does.
The reason most creators don’t run this kind of automation is the visual generation step. AI carousel makers usually output one design style and the slides feel generic. The fix is using a template library where the AI fills in copy against a fixed visual layout, so every output looks intentional instead of stock.
How to Automate Instagram Carousels: The Stack
I use the same four tools across both the Make.com and n8n versions:
- OpenAI to turn a single industry prompt into structured carousel content. You can dump messy business context in and it figures out the slide structure.
- Blotato as the carousel generator. Blotato has a viral template library you point a prompt at, and it maps your content into a finished multi-slide design. This is the part that replaces hand-designing slides in Canva.
- Airtable as the staging table. Every generated carousel lands in a row with the industry tag, the slide images attached, and the source workflow noted. You skim outputs visually before anything ships.
- Make.com or n8n as the orchestrator. Pick whichever you already have a paid seat on. Both work.
I’m involved with Blotato as a creator and tester, so take this with whatever grain of salt feels right. I use Blotato in this stack because the carousel template library is the part that turns generic AI slop into something that looks designed, and because the same workflow can push the finished carousel to Instagram, TikTok, Pinterest, Facebook, or LinkedIn with one extra node.
The carousel template library is the piece that breaks most DIY builds, and it’s why every workflow in this post hands the visual generation step to Blotato. If you want to test the full stack on your own account, start a free 7-day Blotato trial, connect Instagram, and run your first AI-generated carousel end to end. The trial includes the carousel template library, the prompt-to-design API, and enough generation credits to test all three industry branches in this post.
Native Tool vs Schedule-Only SaaS vs AI Workflow: Pick the Right Build
Before you import a blueprint, decide if you actually need this level of automation. If you only post one carousel a week and you enjoy designing in Canva, a scheduler is fine. If you want fresh carousels generating themselves across multiple niches or client accounts, the AI workflow earns its keep fast.
| Approach | What it does | What it can’t do | Best for |
|---|---|---|---|
| Canva + native Instagram | Manual design + manual post | No automation, no batch generation | One creator, under 2 carousels a week |
| Buffer / Later / Plann | Scheduling and analytics for carousels you already designed | Doesn’t generate slide copy or visuals | Solo brands managing 1-3 accounts with a designer on payroll |
| Make.com or n8n + OpenAI + Blotato | Generates slide copy, designs slides from a template library, queues to Airtable, posts | Can’t replace strategy or a strong hook | Creators, agencies, or small businesses running 5+ carousels a week or multiple accounts |
The AI workflow wins on volume, multi-account use, and consistency. The native and scheduling tools win if every carousel needs heavy custom design.
Step 1: Import the Make.com Blueprint and Pick Your Industry
Open the Make.com blueprint from the YouTube description, click the three-dot menu, and hit Import Blueprint. The scenario shows three parallel rows: one for real estate, one for life coaching, one for restaurants. Each row uses the same five-module pattern, just with different prompts and templates wired in.
To control which row runs, edit the Set Variables module at the start. Change the “type” variable to Real Estate, Life Coach, or Restaurant and the router downstream sends the run to the matching branch.

The three keywords are listed in a note on the canvas, so you copy and paste rather than typing and risking a typo that breaks the router. Default is Restaurant, but the real-estate row is the easiest to read end to end on a first run.
Step 2: Run the Real Estate Branch and Watch the Output Build
Hit Run Once with the variable set to Real Estate. The scenario routes to the top branch, calls Blotato to create the visual using a real-estate-focused template, and then hits a sleep module set to 60 seconds. The sleep is there because carousel generation runs async on Blotato’s side and the next module needs the result to be ready before it can pull it down.
Once the sleep finishes, an iterator and aggregator pair stitches the generated slide images into a single record so they all land on one Airtable row instead of getting spread across three or four rows. This is the part that looks like overengineering on the canvas. It isn’t. Without the iterator/aggregator combo, you end up with a messy Airtable view where every slide is its own record.

The final slide of the real estate carousel is a clean call-to-action design with the agent’s brand name and tagline. The whole sequence took about 90 seconds from clicking Run Once to seeing the finished slides in Airtable.
Step 3: Add the OpenAI Prompt-Writer for the Life Coach Branch
The first branch was explicit: slide one says X, slide two says Y, and so on. That works if you already know what you want to say. The life coach branch shows the hands-off pattern, where you delegate slide planning to OpenAI and let Blotato fill in the design.
Open the OpenAI module in the life-coach row. The system prompt says: generate a prompt for an Instagram carousel about an important mindset shift. Output the prompt only, no extra commentary. The user message adds business context: “For context, a life push that helps founders and business owners scale their business to the next level.”

You can give it as much or as little detail as you want about your business, audience, or angle. OpenAI returns a fully-formed carousel prompt, the kind you’d never get around to writing yourself, and that prompt becomes the input to the next Blotato module. Blotato then takes the prompt and intelligently maps the content to the slide template you picked.
The output for this run was a steampunk-style infographic on five mindset shifts for steady growth. The template completely changes the visual feel from the real estate branch.

This is the part of the workflow most “AI carousel” guides skip. You don’t have to micromanage slide-by-slide. You give the system context about your business and it writes the slides for you.
Step 4: Customize the Restaurant Branch With a Brand Image
Set the variable to Restaurant and run it again. The restaurant branch uses a different template: a minimal slideshow with a single centered text block per slide. Output for the run was a five-slide sequence ending on “Reserve your table tonight” with a clean black-on-white type treatment.

The restaurant branch also shows where you can plug in your own brand assets. The Blotato Create Visual module accepts a profile image and an optional custom image input, which is how you’d drop in your actual restaurant logo or hero photo instead of letting the AI generate one. For a small business with existing brand assets, this is the difference between a generic AI carousel and one that looks like part of your actual feed.
Step 5: Rebuild the Same Workflow in n8n
If you prefer n8n, the same three-branch logic ports cleanly with one improvement: the polling pattern. In Make.com you hardcoded a 60-second sleep and hoped Blotato finished in time. In n8n you can build a proper wait loop.

The wait loop starts with a 5-second wait, then hits the Get Visual node to check whether the carousel is ready. If yes, it writes the record to Airtable. If no, it waits another 5 seconds and loops back. The flow keeps polling until the asset is done, then continues. For carousels this typically resolves in 2 to 3 minutes. For longer assets like AI videos this pattern matters even more since generation can take 5 to 10 minutes.
The Blotato Create Visual node inside n8n looks almost identical to the Make.com module: you pick a template from a dropdown, paste your prompt, and the platform handles the rest.

To switch branches in n8n, edit the Set Industry node at the top, drop in Real Estate, Life Coach, or Restaurant, and hit the manual run trigger. The switch routes the run down the correct branch the same way the Make.com router does.
Pro Tips From the Build
Don’t over-engineer the prompt. Kevin’s specific advice from the video: give it the rough idea, give it the business context, and let Blotato handle the heavy lifting. Spending an hour crafting a perfect prompt gets worse output than 30 seconds of “five mindset shifts for founders who want to scale, steampunk style, motivational tone.”
The iterator/aggregator combo in Make.com is the part new builders skip. If your Airtable view ends up with each slide on its own row, that’s why. Add the iterator after the carousel generation node and the aggregator before the Airtable create record node. Slides land together, one row per carousel.
Use the n8n wait loop pattern for anything beyond carousels. AI videos, voiceovers, longer generations all benefit from the polling pattern. A hardcoded sleep guesses at the duration. A wait loop adjusts to it.
What This Workflow Can’t Do (Yet)
This build generates and stages carousels, but a few limits are worth naming up front. Blotato’s carousel API is text-and-template driven, so it doesn’t replace heavily custom designs with multiple image layers and bespoke type treatments. If your brand requires custom illustrations on every slide, you’ll still want a designer.
The Airtable staging step is intentional. You should skim outputs before they ship. AI image generators occasionally render text with typos, and a carousel with a misspelled headline on slide one is the kind of post you regret. Plan to spend 30 seconds glancing at each row before approving it for posting.
Also, the workflows above stop at Airtable. To actually post the carousel to social, you add a Blotato Create Post node at the end of the flow and it pushes to Instagram, TikTok, Pinterest, Facebook, or LinkedIn. Kevin doesn’t walk through that step in this video because it’s covered in separate tutorials, but it’s a single extra node.
Results You Can Expect
A first-time setup for one branch takes about 20 minutes if you already have Blotato, OpenAI, and Airtable connected. After that, generating a new carousel is one click. The full three-industry blueprint runs all three branches in about 6 to 8 minutes end to end. For agencies running multiple client accounts, the per-carousel marginal cost (in OpenAI tokens plus Blotato credits) sits well under what a designer would charge for a single set of slides.
Sabrina’s Final Take
If you post carousels even once a week, this stack pays for itself fast, especially if you run multiple accounts or want to test ten variations of the same hook in a day. Start with one branch (whichever industry is closest to yours), prove the workflow generates something worth shipping, then duplicate the branch for the next niche. The biggest unlock isn’t speed, it’s that you stop avoiding carousels because the design step felt like a chore. For pricing on multi-account use, the Creator and Agency plans on Blotato’s pricing page cover most setups I see in my own community.
Automate Instagram Carousels FAQs
Can you actually automate Instagram carousels end to end?
Yes. The combination of OpenAI for slide copy, Blotato’s carousel template library for visual generation, and Make.com or n8n as the orchestrator gets you from “idea” to “finished multi-slide post” without manual design work. The only manual step you should keep is a quick review in Airtable before publishing.
What’s the best AI model for Instagram carousel slides?
For copy, GPT-4o or GPT-4 Turbo via the OpenAI module is the default. For the visuals, you don’t pick a model directly. Blotato’s template library handles image generation and layout under the hood, which is why this stack produces consistent design across slides instead of the disjointed look you get from feeding raw prompts to a generic image model.
Do I need n8n or Make.com to run this?
Either works. Make.com has a slightly cleaner visual canvas and is easier for first-time automation builders. n8n is more flexible and runs proper polling loops, which matters for longer AI generations. Pick the one you already have a seat on. The video shows both blueprints so you can compare side by side.
How long does one carousel take to generate?
About 60 to 90 seconds for the Blotato step itself. Total run time including the OpenAI prompt and the Airtable write is roughly 2 minutes in Make.com (because of the hardcoded sleep) and 90 seconds to 2 minutes in n8n (because the wait loop exits as soon as the asset is ready).
Why use Blotato instead of going straight to the Instagram API?
The Instagram Graph API requires app review for carousel publishing and rate-limits hard on small accounts. Building it yourself takes weeks. Blotato handles the API plumbing, the rate limits, and the multi-platform posting layer, so the same workflow can also push to TikTok slideshows, Pinterest, Facebook, or LinkedIn with one node change.