Back to Blog
TikTok

How to Get TikTok Photo Posts (Carousels) by API

November 20, 2026
8 min read
S
By SociaVault Team
TikTokTikTok APIPhoto ModeCarouselsWeb Scraping

If you've pulled TikTok data for a while, you've probably hit a "video" with a duration of zero. That's a photo post: TikTok's swipeable carousel of still images with a sound on top. They're everywhere now, from braiding tutorials to NASA's rocket updates, and plenty of scrapers treat them as broken videos.

The short version: a photo post comes back from the API like any other post, but with aweme_type set to 150 and its pictures under image_post_info.images. Here's how to spot one, pull every image, and what you can't get.

What a photo post looks like in the response

We pulled one of NASA's recent carousels, five images of the Artemis III core stage being stacked, through the video-info endpoint. Photo post URLs use /photo/ instead of /video/, and the endpoint accepts either form for the same post:

curl "https://api.sociavault.com/v1/scrape/tiktok/video-info?url=https://www.tiktok.com/@nasa/photo/7701866987609197837" \
  -H "x-api-key: $SOCIAVAULT_KEY"

Trimmed to the fields that matter, here's what came back on November 17, 2026:

{
  "success": true,
  "data": {
    "aweme_detail": {
      "aweme_id": "7701866987609197837",
      "aweme_type": 150,
      "desc": "The core stage of our Artemis III rocket is officially stacked! ...",
      "image_post_info": {
        "title": "Some assembly required 🚀",
        "images": {
          "0": {
            "display_image": {
              "width": 1080,
              "height": 1350,
              "url_list": { "0": "https://...tiktokcdn-us.com/..." }
            },
            "thumbnail": {
              "width": 1206,
              "height": 1508,
              "url_list": { "0": "https://...tiktokcdn-us.com/..." }
            }
          }
        }
      },
      "music": { "title": "City Nights - Instrumental", "author": "DRYVE" },
      "video": { "duration": 0 },
      "statistics": {
        "play_count": 208190,
        "digg_count": 17535,
        "comment_count": 452,
        "share_count": 752,
        "collect_count": 1810
      }
    },
    "transcript": null
  },
  "credits_used": 1,
  "endpoint": "tiktok/video-info"
}

Three things give a photo post away:

  • aweme_type is 150. Regular videos usually report 0.
  • image_post_info.images holds the pictures, one entry per slide.
  • video.duration is 0. The video object is still there, which is why naive code thinks it found a video.

The caption is in desc as usual, and image_post_info.title is the separate title TikTok lets photo posts carry ("Some assembly required 🚀" here). The sound is a normal music object, and the stats work like any post: plays, likes, comments, shares and saves (collect_count).

Each image comes in four versions. display_image is the one TikTok uses for display, 1080 by 1350 for NASA's 4:5 slides. thumbnail was a little larger in our sample (1206 by 1508). The other two, owner_watermark_image and user_watermark_image, are, as their names suggest, versions with TikTok's watermark.

One detail that trips people up: lists in our responses, like images and url_list, can arrive as objects keyed "0", "1" and so on instead of arrays. Read them with a helper that handles both, as below.

Pull every image from a photo post

const API = "https://api.sociavault.com/v1/scrape";
const headers = { "x-api-key": process.env.SOCIAVAULT_KEY };

// Lists can arrive as arrays or as objects keyed "0", "1", ...
const asList = (v) => (Array.isArray(v) ? v : Object.values(v ?? {}));

async function getTikTokPost(url) {
  const res = await fetch(
    `${API}/tiktok/video-info?url=${encodeURIComponent(url)}`,
    { headers },
  );
  if (!res.ok) throw new Error(`HTTP ${res.status}`);
  return (await res.json()).data?.aweme_detail ?? null;
}

function photoImages(post) {
  if (post?.aweme_type !== 150) return []; // not a photo post
  return asList(post.image_post_info?.images).map((image, i) => {
    const img = image.display_image ?? image.thumbnail ?? {};
    return {
      slide: i + 1,
      url: asList(img.url_list)[0] ?? null,
      width: img.width ?? null,
      height: img.height ?? null,
    };
  });
}

const post = await getTikTokPost(
  "https://www.tiktok.com/@nasa/photo/7701866987609197837",
);
const images = photoImages(post);
console.log(images.length); // 5
console.log(images[0].width, images[0].height); // 1080 1350

The same thing in Python:

import os
import requests

API = "https://api.sociavault.com/v1/scrape"
HEADERS = {"x-api-key": os.environ["SOCIAVAULT_KEY"]}

def as_list(value):
    # Lists can arrive as lists or as dicts keyed "0", "1", ...
    if isinstance(value, list):
        return value
    return list((value or {}).values())

def get_tiktok_post(url):
    res = requests.get(f"{API}/tiktok/video-info",
                       params={"url": url}, headers=HEADERS, timeout=60)
    res.raise_for_status()
    return (res.json().get("data") or {}).get("aweme_detail")

def photo_images(post):
    if not post or post.get("aweme_type") != 150:
        return []
    images = []
    for i, image in enumerate(as_list((post.get("image_post_info") or {}).get("images"))):
        img = image.get("display_image") or image.get("thumbnail") or {}
        urls = as_list(img.get("url_list"))
        images.append({"slide": i + 1, "url": urls[0] if urls else None,
                       "width": img.get("width"), "height": img.get("height")})
    return images

post = get_tiktok_post("https://www.tiktok.com/@nasa/photo/7701866987609197837")
print(len(photo_images(post)))  # 5

One call costs 1 credit, whether the carousel has 2 slides or 35.

Find every photo post from a creator

Photo posts are mixed into the same feed as videos, so there's no separate endpoint. Page through a creator's posts and keep the ones with aweme_type 150. On the first page of NASA's feed, 1 of the 10 posts was a carousel.

async function photoPostsFromCreator(handle, maxPages = 5) {
  const found = [];
  let cursor;
  for (let page = 0; page < maxPages; page++) {
    const params = new URLSearchParams({ handle });
    if (cursor) params.set("max_cursor", cursor);
    const res = await fetch(`${API}/tiktok/videos?${params}`, { headers });
    if (!res.ok) break;
    const data = (await res.json()).data ?? {};
    for (const post of asList(data.aweme_list)) {
      if (post.aweme_type === 150) {
        found.push({
          id: post.aweme_id,
          slides: asList(post.image_post_info?.images).length,
          plays: post.statistics?.play_count ?? null,
        });
      }
    }
    if (!data.has_more || !data.max_cursor) break;
    cursor = data.max_cursor;
    await new Promise((r) => setTimeout(r, 500)); // be gentle between pages
  }
  return found;
}

console.log(await photoPostsFromCreator("nasa", 3));

Each page costs 1 credit and returns about 10 posts, and maxPages caps the loop so a busy account can't run up your bill. Feed items carry the same image_post_info as video-info, so you don't need a second call per post.

Comments work the same way on photo posts. Pass the /photo/ URL to the comments endpoint just as you would a video URL.

Every image URL is a signed CDN link with an expiry baked in. The links in our sample were set to expire about 15 days after we fetched them. We measured how long TikTok, Instagram and Facebook media links last in our study of 460,000 expiry timestamps; the short version is: don't store the URL and expect it to work next month. Download the files you need, or fetch the post again when you need fresh links.

The images belong to the creator. Analyzing them is fine; republishing them is a question for the creator, not for an API.

What this can't do

  • No transcript. A carousel has no speech to transcribe; the sound is usually a music track. Video-info returns transcript: null for photo posts, and the transcript endpoint returns an error saying the post is a photo carousel, without charging a credit.
  • No text from inside the images. TikTok tags each image with an ocr_language, a hint that it reads the text on the slides, but that text isn't in the response. If you need it, run OCR on the images you download.
  • No per-slide stats. Plays, likes, comments and saves cover the whole post. There's no way to see which slide people stopped on.
  • No private accounts. Like any private content, photo posts from private accounts aren't available.

Frequently Asked Questions

Can you get TikTok photo posts through an API?

Yes. Photo posts come through the same endpoints as videos: video-info for a single post, and the videos endpoint for a creator's feed. Look for aweme_type: 150 and read the images from image_post_info.images.

How do I tell a TikTok photo post from a video?

Check aweme_type. Photo posts report 150, and regular videos usually report 0. A photo post also has image_post_info.images filled in and a video.duration of 0.

Yes. Video-info and comments accept /photo/ URLs, /video/ URLs for the same post, and TikTok short links. See how to turn short links into the real post if you're working from shared links.

How many credits does a photo post cost?

One credit per video-info call, however many images the carousel has. Pulling a creator's feed costs 1 credit per page of about 10 posts.

Do TikTok photo post image URLs expire?

Yes. They're signed CDN links, and the ones in our sample expired about 15 days after we fetched them. Download the images you need, or fetch the post again later for fresh links.

Can I get a transcript or the text inside the images?

No transcript, because a carousel has no speech; the transcript endpoint returns an error for photo posts and doesn't charge for it. The text inside the images isn't returned either, so run OCR yourself if you need it.


Want to try it on a carousel you care about? Grab 50 free credits, no card needed. For the rest of the post fields, see the TikTok video info API guide, and if you also track Instagram carousels, our Instagram carousel analytics guide covers the other side.

Found this helpful?

Share it with others who might benefit

Ready to Try SociaVault?

Start extracting social media data with our powerful API. No credit card required.