AI Carousel Generator via MCP: Create Instagram Carousels in Minutes, Not Hours

For a long time, making a carousel meant opening Figma, copying yesterday's frames, swapping the text, exporting ten PNGs, dragging them into Instagram, and then spotting a typo on slide 4. Each step was small, but they added up. Most solo operators we talk to spent about 90 minutes per carousel before they automated it. AI can now generate the slides, but that's not the interesting part. With MCP, you can generate the carousel inside the tool you're already working in.

Anthropic's Model Context Protocol turned a small detail into a meaningful shift. Instead of switching to a separate carousel tool, you can ask Claude (or any MCP-aware client) to draft a 7-slide carousel from a one-line idea, render it with your brand fonts, and hand back a preview URL — all without leaving the chat. The work doesn't get faster because AI is better. It gets faster because the friction between "thinking about the post" and "posting it" collapses.

Key takeaways

• MCP is an open standard for connecting AI clients to outside tools. The current spec is from November 2025.
• The Mirr MCP server lets Claude generate, refine, and render carousels by calling tools in the middle of a conversation.
• In Q1 2026, Mirr's carousel pipeline took 43 seconds on average from start to finish, down from 4 min 12 sec in mid-2024.
• The speed came from cutting out handoffs between tools, not from a smarter model. Deciding what's worth posting is still your job.

What MCP actually does for carousels

MCP is a standard that lets AI clients talk to outside tools in a consistent way. For carousels, it's simple: a server offers a few tools — carousel_generate, carousel_refine_page, carousel_render_images — and Claude can call them with structured inputs while the conversation keeps going.

This matters because the old workflow kept breaking your train of thought. You'd describe the post in one tool, paste the result into another, edit it there, then schedule it somewhere else. Every handoff was a chance to lose track of what you were trying to say. With MCP, the carousel takes shape in the same conversation where the idea started.

A typical run, step by step

You open Claude (or Claude Code, or Cursor, or any MCP client) with the Mirr MCP server connected. You type something like "draft a 7-slide carousel about why Instagram down-ranks engagement bait, in our voice." About 12 seconds later, you get a JSON structure with 7 pages. Each page has a headline, body text, and a layout suggestion. You review the copy in the chat and ask for a sharper slide 3. The model edits it right there.

Then you ask for the visual version. The server sends back a preview link with all seven slides laid out in your brand fonts and colors. If a slide looks crowded, you say so. If a font looks wrong, you swap it. All of this happens in the chat, and the finished carousel is saved in your account, ready to schedule.

The one number worth quoting: in Q1 2026, the average time on Mirr's carousel pipeline was 43 seconds from prompt to rendered preview. The first version, in mid-2024, took 4 minutes 12 seconds. The speed didn't come from a better model. It came from cutting out handoffs between tools.

Where this stops being magic

Generating carousels through MCP has a few real limits. Claude is good at structuring copy, but it still sometimes writes too much, with long captions where short ones would hit harder. Fix this in the prompt itself. Give it clear limits ("each slide max 14 words"), and the output gets noticeably tighter. Don't count on the model to edit itself.

The second limit is taste. AI can make a carousel that's well laid out and on-brand, but it can't tell you whether the post is interesting. That part still has to come from you. The tool takes care of the production busywork. It doesn't do the editorial thinking for you.

And finally, MCP servers are still early. The spec hit a stable November 2025 release, but the ecosystem of clients and servers is younger than it looks. Expect occasional version drift between client and server, especially on the sampling side. The protocol is maturing fast, but it's not yet boringly reliable across all clients.

Why this matters more than another generator

The web is full of AI carousel generators now. Most are the same two image models with different templates on top. The MCP version differs in one small but important way: it works wherever you already do your thinking.

If your daily work happens in Claude or Claude Code, the carousel can come out of that work instead of being a separate trip. We built the Mirr MCP server around this idea. The goal isn't to be a better generator. It's to let the carousel happen inside the conversation where the idea came from, with no separate step.

If your team doesn't want to set up its own MCP connection, the same process is available in Mirr Carousel Lab: start with a topic, generate the slide structure, edit the copy, render the images, and schedule the post from the same workspace.

Good tools get out of your way. For a lot of small marketing teams, carousels were still in the way. MCP finally moves them aside.

Frequently asked questions

What MCP clients work with Mirr's carousel server?

Claude Desktop, Claude Code, Cursor, and Zed all work as of April 2026. Any client that supports the MCP tool-calling spec at version 0.6 or later can connect.

Do I need a Mirr paid plan to use the MCP server?

The free plan includes 5 carousel generations per day through MCP. For more, you'll need a paid plan, mainly because rendering takes computing power that isn't free.

Can the AI use my own brand fonts and colors?

Yes. The server pulls from your Mirr workspace's brand kit, so the rendered output uses whatever fonts, colors, and layout presets you've already saved.

Is the carousel editable after generation?

Yes, in two ways. You can keep refining it through chat using carousel_refine_page, or open it in Mirr's web editor and change it directly. Both versions stay in sync.

Mirr is a social marketing tool for solo SaaS builders and small teams. Carousels, card news, and scheduled publishing in one place. Connect Mirr MCP →

Want to make these without a designer? Try Mirr's AI carousel generator — build a full carousel from one idea, free.