How Many Slides Should a TikTok Slideshow Have? Our Data

Slide count is one of the first decisions in our complete guide to TikTok slideshows, and one of the most over-advised. Here's what our own analysis data says.

How many slides should a TikTok slideshow have?

TikTok lets you post up to 35 images in one slideshow, but there's no proven best number. In the slideshows we analysed, the median was 3 slides and 2 was the most common length, and slide count showed no relationship with views relative to followers. Use as many slides as the idea needs, and put the payoff on the last or next-to-last slide.

That's the honest answer, and it's less tidy than the "5 to 10 slides" you'll see elsewhere. Below: what the 35-image limit really is, what length the slideshows in our sample actually used, why more slides didn't mean more views, and how to pick a length from the structure of your idea.

The limit: 35 images

TikTok's own developer documentation for photo posts describes the image list as "an array containing up to 35 photo content URLs". That's the cap for posting through the API, and Photo Mode in the app is widely reported to allow the same 35 (TikTok's Photo Mode announcement from October 2022 doesn't state a number, so check the app if you're planning a very long one).

Almost nobody gets near it. The longest analysed slideshow in our sample had 20 slides. In our niche catalogue, a separate set of photo posts pulled from the feeds of top creators in each niche, the longest reached 35, but the median was 3.

What length the slideshows we analysed used

FYPNow reads TikTok slideshows slide by slide: the cover, each slide's text and role, where the payoff lands, and our model's rating of how strong the cover is and how likely people are to swipe through and save. Here's how long the slideshows we've analysed were.

The median is 3 slides. Nearly half the posts had 2 or fewer, about a quarter had 5 or more, and fewer than 1 in 10 went past 8.

Read this before quoting it. These slideshows come from accounts that FYPNow customers chose to track or inspect, posted between March 2024 and October 2026. That's a convenience sample, not a random sample of TikTok:

• A few creators account for a large share of the sample. The most prolific of them has a median of 2 slides (just over half their posts are two-slide). Take them out and the median is still 3, but the two-slide spike shrinks from 35% to 27%.
• The sample leans toward business, finance, motivation and education. A recipe or travel account may look quite different.
• Every role, label and score comes from one AI model reading the slides. They're our model's ratings, not measured behaviour.

So "most slideshows are short" holds for this sample. It isn't a law of TikTok.

Do more slides get more views?

No, not in our data. We compared slide count with views divided by the creator's follower count (only where the creator had at least 1,000 followers, so the ratio means something). The Spearman rank correlation was −0.03. That's flat: no relationship.

Raw views show a faint positive link (rho 0.13), but it doesn't survive a closer look. The view medians per length mostly reflect which creators sit in each bucket. The two-slide bucket, for example, comes from only a handful of creators. One big account in a bucket moves its median more than slide count does.

That's why this post doesn't print a "views by slide count" table. The numbers exist, but they'd tell you about a handful of accounts, not about slide count.

What longer slideshows did get: higher model ratings

Our model did rate longer slideshows as more swipe-worthy. The rank correlation between slide count and the model's swipe-through score was 0.34, and it held at 0.34 without the largest creator. Save-worthiness ratings rose more weakly with length (rho 0.23).

Ratings are 0–100, from our model. Two things keep this from being advice to add slides:

• Part of the link is built in. A one-slide post gets a swipe rating near 10 because there's nothing to swipe.
• Past that, the ratings bounce around rather than climb: 85 at two slides, 65 at three, 62 at six to seven, 85 again at eight to ten. A score that goes up and down like that is following content, not length.

A model rating also isn't what TikTok measures. TikTok doesn't expose per-slide swipe-away data for other people's posts, so swipe-through on someone else's slideshow can't be measured from outside.

Why the "5 to 10 slides" advice is shaky

Several slideshow guides lead with 5 to 10 slides (Supaslides, for example, attributes the range to WaveGen). We couldn't find a published dataset behind the range, and our own sample doesn't support any band. The median slideshow we analysed had 3 slides, and length told us nothing about how a post did against its creator's follower count.

The reasoning behind the advice is fair, though: a deck people finish is better than one they abandon. That's an argument for no wasted slides, not for a particular count.

How to choose your slide count: work back from the payoff

Instead of a number, use the structure. In the slideshows we analysed, our model labelled slide 1 as the hook in 98% of them, and placed the payoff at or near the end. In slideshows with 4 or more slides, the payoff sat on the last or next-to-last slide in 83% of them, and on the very last slide in 50%. Where the last slide wasn't the payoff, it was most often a call to action: 20.4% of the multi-slide posts ended on a CTA slide.

So the shape is fixed and the middle stretches. Your slide count is:

1 cover + however many slides the middle really needs + 1 payoff (+ 1 optional CTA).

A few worked examples:

• One sharp point with a reveal. "This is the setting nobody turns on" → the setting. Two slides. This was the single most common shape in our sample (31% were simply hook then payoff).
• A short list. "3 mistakes first-time gardeners make" → three mistakes → the fix. Five slides. If you promise 3, you need at least 3 middle slides; the number on the cover sets your length.
• A tutorial. Cover with the result → one slide per step → the finished thing. Let the steps decide, and merge steps that don't need their own picture.
• A story. Hook → setup → the turn → the outcome. Four or five slides, and the payoff is the outcome.

Then cut. For each middle slide, ask whether someone who skipped it would miss anything. If not, merge it or drop it. The slideshow ideas guide gives a slide-by-slide sequence for each common structure, and the step-by-step guide to making a TikTok slideshow covers adding and reordering slides in the app.

When one slide is enough

15% of the slideshows we analysed were single images. A single slide works when the whole idea fits on the cover: a quote, a meme, a striking photo. It gives up the swipe entirely, so don't add slides to a one-slide idea just to make it a carousel.

When to go long

Long decks earn their length when every slide is a reference item someone will come back to: a full checklist, a recipe with each step, a set of templates. The most common save reason our model gave across the slideshows we analysed was "something to come back to" (48% of posts), and that kind of content is naturally longer. But that's a hypothesis from our model, not observed saves.

Check your own slideshows

Your account's numbers beat any sample. Pull your last ten slideshows, note each one's length and views divided by your followers, and see whether length lines up with anything. If you analyse a slideshow in FYPNow's Video Analysis, you get the slide-by-slide roles and where the payoff lands, so you can see whether a long post had a slow middle or a short one ran out of room. The slideshow docs explain what the breakdown includes.