TikTok Video Frame Extractor - Video to JPG, PNG & WebP

Turn any TikTok video into still images. Paste video links or usernames, choose how frames are picked (every N seconds, evenly spaced, at every scene change, keyframes or exact timestamps) and get clean, watermark-free JPG, PNG or WebP frames with permanent download links. No login, no TikTok account and no software to install.

What you get

• Watermark-free frames cut from TikTok's standard H.264 video (720x1280 for portrait videos), or scaled down to any width you choose
• Five extraction modes: evenly spaced count, fixed interval, scene changes, keyframes (I-frames) and exact timestamps
• JPG, PNG or WebP output, plus an optional contact sheet: one grid image of every frame stamped with its timestamp
• Whole profiles: enter a username and the latest videos are processed automatically
• Photo slideshows: slideshow posts return their original full-size photos instead of frames
• For every frame: frameUrl, timestamp, timestampSec, sceneScore, width, height, imageFormat, kvKey
• For every post: creator author and authorName, description (caption), createdAt, views, likes, comments, shares, durationSec, framesExtracted, frameUrls and contactSheetUrl
• Export to JSON, CSV, Excel or Google Sheets, or pull the images straight from the API

Use cases

• Thumbnail and cover creation: grab the best-looking frame of a TikTok for a blog post, YouTube thumbnail or ad
• Content analysis and computer vision: feed frames into image classifiers, OCR, product or logo detection
• Competitor and trend research: build storyboards of a creator's latest videos to study hooks, scenes and editing pace
• Brand monitoring and compliance: capture visual proof of how a product or logo appears in creator content
• Machine learning datasets: collect evenly spaced frames at scale for training and labelling

How to use

1. Paste TikTok video links into TikTok video links and/or usernames (with or without @) into TikTok profiles
2. For profiles, set Videos per profile (default 10 latest posts)
3. Pick an Extraction mode:
   • Count: N frames spread evenly across the video (default 8)
   • Interval: one frame every N seconds (default 2)
   • Scenes: one frame at each shot change, tuned with the sensitivity threshold
   • Keyframes: the video's own I-frames
   • Timestamps: exact times such as 3, 0:12 or 15.5s
4. Choose Image format and, optionally, a Max frame width and Contact sheet
5. Run the actor. Frames appear in the Dataset tab with image previews, and the files are in the Key-value store tab

Use Start time and End time to only look at part of each video, and Max frames per video to cap the output.

Output format

One row per frame:

{
  "recordType": "frame",
  "status": "ok",
  "postType": "video",
  "videoId": "7655004069190192398",
  "videoUrl": "https://www.tiktok.com/@zachking/video/7655004069190192398",
  "author": "zachking",
  "authorName": "Zach King",
  "description": "a message to all AI",
  "createdAt": "2026-06-24T16:45:49.000Z",
  "views": 67800000,
  "likes": 14000000,
  "mode": "count",
  "frameIndex": 1,
  "timestampSec": 2.329,
  "timestamp": "00:00:02.329",
  "width": 720,
  "height": 1280,
  "imageFormat": "jpg",
  "kvKey": "zachking-7655004069190192398-frame-001.jpg",
  "frameUrl": "https://api.apify.com/v2/key-value-stores/.../records/zachking-7655004069190192398-frame-001.jpg",
  "durationSec": 27.95
}

Each post also gets one summary row (recordType: "video") with framesExtracted, frameUrls, contactSheetUrl and the minutes billed. Slideshow photos use recordType: "photo". Links that could not be processed (deleted or private videos, unknown usernames, invalid links) get a recordType: "notice" row explaining why, and are never charged.

Pricing

You pay per result, and only for posts that returned at least one image:

Event	Price	When
Video processed	$0.02	Once per video or slideshow. Includes the first minute of video
Frame	$0.0007	Per video frame saved
Photo	$0.0025	Per original slideshow photo saved
Video minute	$0.006	Per started minute of video length after the first minute, in every mode (most TikToks are under a minute and never pay this)
Scene-scan minute	$0.004	Scenes mode only: per started minute of video scanned for shot changes

Example: 8 frames from a 30-second TikTok cost $0.0256. Scene detection on the same video with 10 scenes found costs $0.031.

Failed, private, deleted and skipped posts are free. The contact sheet is included. The summary row of each video shows the minutes billed (billedVideoMinutes, billedScanMinutes). You can set a maximum cost per run in the run options and the actor stops cleanly when it is reached.

Tips

• Scene detection works best on edited videos with clear cuts. Lower the threshold to 0.2 to catch soft transitions, or raise it to 0.5 for hard cuts only.
• Pinned posts are skipped by default when you enter a profile, since they are often old. Turn on Include pinned posts to keep them.
• Free plan runs are limited to 5 posts per run. Upgrade to a paid Apify plan to remove the limit.

FAQ

Are the frames watermarked? No. Frames are cut from the original, watermark-free video.

Can I extract frames from private videos? No. Only public videos can be processed.

How long are the download links valid? The images live in the run's key-value store, so they stay available for as long as Apify keeps the run's storage (according to your plan's data retention).