# Video Frame Extractor: Keyframes & Thumbnails (`enisbodlli/video-keyframe-extractor`) Actor Extract frames from video files, uploads and video podcast feeds: a set number, one every N seconds, scene changes, keyframes or listed times, plus a contact sheet. Skips blank and repeated frames. JPG, PNG or WebP with links. Pay per frame stored. - **URL**: https://apify.com/enisbodlli/video-keyframe-extractor.md - **Developed by:** [Enis Bodlli](https://apify.com/enisbodlli) (community) - **Categories:** Videos, AI, Developer tools - **Stats:** 2 total users, 1 monthly users, 100.0% runs succeeded, 0 bookmarks - **User rating**: No ratings yet ## Pricing from $1.50 / 1,000 frames This Actor is paid per event. You are not charged for the Apify platform usage, but only a fixed price for specific events. Since this Actor supports Apify Store discounts, the price gets lower the higher subscription plan you have. Learn more: https://docs.apify.com/actors/running/actors-in-store.md#pay-per-event ## What's an Apify Actor? An Actor is a serverless cloud program that runs on the Apify platform. It has two run modes. In Batch mode, an Actor accepts a well-defined JSON input, performs an action which can take anything from a few seconds to a few hours, and optionally produces a well-defined JSON output, datasets with results, or files in key-value store. In Standby mode, an Actor provides a web server which can be used as a website, API, or an MCP server. Apify vocabulary and the platform model are defined once, in the agent quickstart at https://apify.com/agents.md. ## How to integrate an Actor? If asked about integration, you help developers integrate Actors into their projects. You adapt to their stack and deliver integrations that are safe, well-documented, and production-ready. Do not guess an integration path. Every one of them is in the agent quickstart at https://apify.com/agents.md: the Apify MCP server, Agent Skills with the Apify CLI, the JavaScript and Python clients, the REST API, and the account-free path for an agent with no human to sign in. It also carries the rule on stating cost before the first paid run. For examples already wired to this Actor's own input schema, see the [API](#api) section below. Each client library has reference documentation the quickstart does not restate: [JavaScript/TypeScript](https://docs.apify.com/api/client/js/docs.md) (`npm install apify-client`) and [Python](https://docs.apify.com/api/client/python/docs.md) (`pip install apify-client`). # README ## Video Frame Extractor: Keyframes & Thumbnails Turn **videos into images**: a set number of **frames**, one every few seconds, a frame at each **scene change**, the video's own **keyframes**, or frames at the **times you list**, plus an optional **contact sheet** that shows the whole video on one image. Give it links to video files, web pages that embed a video, **video podcast RSS feeds**, or upload your own files. Every frame comes back as a JPG, PNG or WebP with a link and its exact timestamp. It is built for people who make thumbnails and previews, teams that feed video into AI vision models, researchers who code footage, and AI agents that need to see what is in a video. - **Only useful frames, and only those are charged.** Black, white and faded frames, and frames that repeat the one before (a slide that stays on screen), are skipped for free. - **Scenes with their scores, or keyframes for speed.** Scene mode gives each cut's strength; keyframe mode reads a long video in a fraction of the time. - **Pay per frame, $2 per 1,000 frames** ($1.50 on Gold), with nothing to pay for failed links, skipped frames or platform usage. The easiest way to try it: keep the example input and press **Start**. You get up to six frames and a contact sheet from each of two short films in about 10 seconds. ### What you get for each frame ```json { "status": "extracted", "type": "frame", "videoUrl": "https://media.w3.org/2010/05/sintel/trailer.mp4", "videoTitle": "trailer.mp4", "mode": "count", "frameIndex": 1, "timestampSeconds": 4.375, "timecode": "00:00:04.375", "imageUrl": "https://api.apify.com/v2/key-value-stores/STORE_ID/records/frame-aed5bc55391c5011-00001.jpg", "imageWidth": 854, "imageHeight": 480, "contactSheetUrl": "https://api.apify.com/v2/key-value-stores/STORE_ID/records/contact-sheet-aed5bc55391c5011.jpg", "durationSeconds": 52.208, "fps": 24, "videoCodec": "h264", "billedCount": 1 } ``` ### What you can do with it - **Thumbnails and preview strips** for a video library, a course or a podcast's video episodes. - **Frames for AI vision models**: send the frames to GPT, Claude or Gemini to describe, tag, moderate or search what is in a video, without sending the whole file. - **Scene and shot lists** for editing, research or archive work, with a timecode per cut. - **Slide capture** from recorded talks and webinars: one image per slide, repeats left out. - **Training data** for computer vision: frames at a fixed rate, at the size and format you need. - **A new video podcast episode, as pictures**: point it at the feed and schedule it. ### How to use it 1. Paste one or more links in **Video links**, or upload files in **Upload video files**. A direct file link (MP4, MOV, WEBM, MKV, AVI, FLV, MPEG-TS and more), a page that embeds a video file, and a video podcast RSS or Atom feed all work. 2. Choose **How frames are chosen**: - `count`: a set number of frames spread evenly over the video (default 10). - `interval`: one frame every few seconds, from 0.1 s upwards. - `scenes`: a frame at each cut or scene change, with its score; lower **Scene change sensitivity** finds softer cuts. - `keyframes`: the frames the video was encoded around, the fastest way through a long video. - `timestamps`: frames at the times you list, as seconds (`90`), clock times (`1:30`) or a share of the video (`50%`). 3. Optionally set **Start at** and **End at** to work on part of each video, a **Maximum width**, the **Image format** and quality, and switch on **Make a contact sheet**. 4. Press **Start**. Frames appear in the dataset as they are stored; open the **Overview** tab to see them as images, or download the dataset as JSON, CSV or Excel. #### How long it takes Measured on Apify at the default 512 MB on 2026-10-20, with a 10-minute 720p film: | What | Images | Time | |---|---|---| | One frame every 10 seconds | 56 | 10 s | | Scene changes, with a contact sheet | 123 | 17 s | | One frame every 0.6 seconds | 784 | 2 min 17 s | Each video is downloaded once, so a large file adds its download time. Videos are processed one after another; more memory gives the run more CPU, which shortens long videos. ### Pricing You pay per image stored, a frame or a contact sheet, plus Apify's run start of $0.00005 per GB of memory ($0.00005 at the default 512 MB). Platform usage is included, so there is nothing else to pay. | Apify plan tier | Per frame | Per 1,000 frames | |---|---|---| | Free, Bronze | $0.002 | $2.00 | | Silver | $0.00176 | $1.76 | | Gold and above | $0.0015 | $1.50 | Worked examples at the Free and Bronze price: - The example input, 11 frames and 2 contact sheets: 13 x $0.002 = **$0.026**. - 1,000 frames: **$2.00**, or $1.76 on Silver and $1.50 on Gold. - One frame every 10 seconds from an hour of video: 360 frames, **$0.72**. - Ten thumbnails and a contact sheet for each of 100 videos: 1,100 images, **$2.20**. - A blank or repeated frame, a link that fails, or a video with no usable frame: nothing. On the Apify free plan, the monthly $5 of credit pays for about 2,500 frames. **Maximum frames per video** caps what each video can cost, and you can set a maximum charge per run: the Actor then stores only as many frames as that amount pays for, keeps a place for the contact sheet, and finishes as succeeded with a message that says the limit stopped it. A run that Apify restarts on another server carries on where it stopped and charges no frame twice. ### Input ```json { "videoUrls": [ "https://media.w3.org/2010/05/sintel/trailer.mp4", "https://feeds.twit.tv/twit_video_hd.xml" ], "maxVideosPerFeed": 1, "mode": "scenes", "sceneThreshold": 0.3, "maxFramesPerVideo": 50, "startTime": "0:30", "imageFormat": "webp", "maxWidth": 1280, "contactSheet": true, "contactSheetColumns": 5 } ``` | Field | What it does | Default | |---|---|---| | `videoUrls` | Links to video files, pages that embed one, or video podcast feeds | example films | | `videoFiles` | Uploaded files, or links to records in an Apify key-value store | none | | `maxVideosPerFeed` | How many of a feed's newest videos are processed | 3 | | `mode` | `count`, `interval`, `scenes`, `keyframes` or `timestamps` | `count` | | `frameCount` | Frames per video in `count` mode | 10 | | `intervalSeconds` | Seconds between frames in `interval` mode | 10 | | `sceneThreshold` | Scene change sensitivity, 0.01 to 0.99 | 0.3 | | `timestamps` | Times for `timestamps` mode | none | | `maxFramesPerVideo` | At most this many frames per video in `interval`, `scenes` and `keyframes` | 100 | | `startTime`, `endTime` | The part of each video to work on | whole video | | `skipBlankFrames` | Leave out black, white and faded frames, free | on | | `skipDuplicateFrames` | Leave out a frame that looks the same as the last one kept, free | on | | `imageFormat` | `jpg`, `png` or `webp` | `jpg` | | `maxWidth` | Scale wider frames down to this width; 0 keeps the video's size | 1280 | | `imageQuality` | JPG and WebP quality, 1 to 100 | 85 | | `contactSheet` | One grid image per video with every frame and its time | off | | `contactSheetColumns` | Frames per row on the contact sheet | 4 | ### Output One dataset row per image: `type` is `frame` or `contact-sheet`. Every row also carries the video's duration, size, frame rate, codec and whether it has sound, and for a feed the feed's and episode's title and date. A link that fails gives one row with `status` `failed` and the reason. ```json [ { "status": "extracted", "type": "frame", "inputUrl": "https://test-videos.co.uk/vids/bigbuckbunny/mp4/h264/360/Big_Buck_Bunny_360_10s_1MB.mp4", "sourceType": "file", "videoIndex": 2, "videoTitle": "Big_Buck_Bunny_360_10s_1MB.mp4", "mode": "count", "frameIndex": 3, "timestampSeconds": 4.167, "timecode": "00:00:04.167", "sceneScore": null, "imageUrl": "https://api.apify.com/v2/key-value-stores/STORE_ID/records/frame-822a3b8b2647449b-00003.jpg", "imageKey": "frame-822a3b8b2647449b-00003.jpg", "imageFormat": "jpg", "imageWidth": 640, "imageHeight": 360, "imageBytes": 37086, "framesInVideo": 6, "durationSeconds": 10, "width": 640, "height": 360, "fps": 30, "videoCodec": "h264", "hasAudio": false, "billedEvent": "frame", "billedCount": 1, "error": null, "errorKind": null }, { "status": "extracted", "type": "contact-sheet", "videoIndex": 2, "videoTitle": "Big_Buck_Bunny_360_10s_1MB.mp4", "imageUrl": "https://api.apify.com/v2/key-value-stores/STORE_ID/records/contact-sheet-822a3b8b2647449b.jpg", "imageWidth": 1308, "imageHeight": 380, "framesInVideo": 6, "billedEvent": "frame", "billedCount": 1 }, { "status": "failed", "type": null, "inputUrl": "https://www.youtube.com/watch?v=abc", "imageUrl": null, "billedCount": 0, "error": "YouTube links are not supported: this Actor takes frames from video files, web pages that embed a video file, and video podcast feeds. Use a direct link to a video file you may use, or upload it in \"videoFiles\".", "errorKind": "unsupported-platform" } ] ``` The images themselves are in the run's key-value store, named `frame-