Skip to main content

Dark-themed desktop tool for trimming videos, downloading clips from the open web, and exporting polished GIFs or MP4s with text overlays.

Project description

Grab a video from the web, cut the part you want, caption it, and export a clean GIF or MP4. Everything runs on your PC. No watermark, no account, no upload.

Animated tour of the app


What it does

  • Download from the web. YouTube, Instagram, X, Reddit, Bluesky, Facebook, and 1,000+ other sites yt-dlp supports. Paste a link, press Ctrl+V from anywhere, or queue a batch. Reads browser cookies for private and age-gated videos.
  • Trim to the exact moment. Frame-accurate timeline with a waveform and thumbnail strip. Queue several cuts from one video, then export them separately or stitched with transitions.
  • Captions that look right. Text with outline, shadow, bold, italic, and multiple lines, and the preview matches the exported frame exactly. Auto-caption speech with Whisper, or import an .srt or .vtt.
  • Captions that follow the action. Pin a caption to a moving subject and it tracks them across the frame, the way the "click for more" memes do. Drag to set the path by hand, or hit ⚡ Auto-track and let it follow the subject for you. Preview and export stay in sync.
  • Overlays. Logos, watermarks, and sticker or GIF overlays dragged anywhere on the frame, each with its own size, opacity, and timing. Paste an image straight from the clipboard.
  • Export for the platform. Tune GIFs (dither, palette, loop) or export hardware-encoded MP4s. Reframe 16:9 to 9:16 for Shorts, Reels, and TikTok with a blurred-background fill. Speed, rotate, mute, audio-only, and colour adjustment are built in.
  • Record your screen straight into the editor.
  • Save real projects. .vidkid project files preserve the source link, trim ranges, crop, captions, overlays, and export choices. Autosave recovery protects work after an interrupted session, and recent projects stay one click away.
  • Update through the right channel. The app detects Store, winget, pip, APT, AppImage, macOS, portable, and source installs, then offers the safest update route for that install.
  • Runs fully offline. No upload, no account, no watermark. Open source, FFmpeg included.

One tab does it all: open a file, record the screen, or paste a link, then trim, caption, and export in the same place. A fixed tool dock keeps every section one click away, even when a project has many caption lines. VideoKidnapper also has undo and redo, project save and recovery, a Batch Export tab, an export History tab, keyboard shortcuts (Ctrl+S project, Space play, J/L step, I/O in-out, Ctrl+E export), a CLI mode, and a plugin API. See docs/ROADMAP.md for what is next.


Tabs & features

Kidnap & Trim

One tab for the whole job. Load a video three ways (Open Video File, Record Screen, or paste a link), then trim, caption, and export in the same place.

Kidnap & Trim with a video loaded, a caption, and a queued range

With a video loaded you get a thumbnail strip, a waveform, a dual-handle range slider, and queued ranges (each exports as its own clip, or they concatenate with transitions when "Concat queued ranges" is on). Captions expose per-layer controls: multiline text, bold, italic, outline, and shadow, and the caption renders on the preview exactly as it will export. Import SRT and 🗣 Auto-captions (Whisper) both feed the same text-layer panel.

Turn on 🎯 Motion track and a caption follows a moving subject across the frame, like the "click for more" memes. Drag the caption at a few points in the clip to set its path by hand, or drop the caption on your subject and hit ⚡ Auto-track to have it follow them automatically (OpenCV object tracking, bundled in the Store, Windows, macOS, and Linux AppImage builds). The same interpolation drives the preview and the export, so what you see is what renders.

One caption tracked across three moments of a clip

The download bar and batch queue

The Kidnap from bar detects the platform as you paste (or press Ctrl+V to drop a link in from anywhere in the app). Supported with brand chips:

Platform Host patterns
YouTube youtube.com, youtu.be, music.youtube.com, m.youtube.com
Instagram instagram.com
Bluesky bsky.app, bsky.social
Twitter/X twitter.com, x.com, mobile.twitter.com
Reddit reddit.com, redd.it, v.redd.it (gallery-wrapped + video+audio auto-merged)
Facebook facebook.com, fb.watch, fb.com, m.facebook.com

…plus the 1,000+ other sites yt-dlp supports. Cookies from reads login cookies from Chrome / Firefox / Edge / Brave / Opera, or a cookies.txt export, for private/age-gated videos. (Windows Chrome encrypts its cookie DB, so close Chrome fully, use Firefox, or a cookies file; the in-app error explains it.) Downloads retry transient failures with resume, and ⟳ Update yt-dlp keeps the extractor current. Batch Download takes a list of links, grabs them in order, and loads any finished one into the editor with Use.

History

History tab

Every successful export is persisted to ~/.videokidnapper_settings.json — the 25 most recent show here with format, quality preset, timestamp, and file size. Open launches the file in the default system player; Reveal opens its folder in Explorer / Finder. Missing files (moved or deleted) are dimmed and their buttons disabled.

Projects and recovery

Project save and recent-file hub

Press Ctrl+S to save a .vidkid project, Ctrl+Shift+S to save a copy, or Ctrl+Shift+O to open one. Project files keep the source video reference, trim and queued ranges, crop, text and image layers, plus export choices. Relative media paths travel with a project folder when possible. Unsaved edits are written atomically to a recovery file, and the next launch offers to recover it after an interrupted session.

App updates

Install-aware update prompt

When a release is available, the header opens an update prompt tailored to the current install. Store and native package-manager installs use their signed or verified update route. Portable, AppImage, macOS, and source installs open the matching verified release instead of replacing a running executable in place.

Debug

Debug tab with color-coded levels

Captures stdout + stderr with level-colored tags: INFO accent-blue, WARN amber, ERROR red. Uncaught exceptions from Tk callbacks and normal Python code both land here via a global exception hook, so a crash leaves a traceback instead of killing the app. Useful when yt-dlp reports a protected video or ffmpeg rejects a filter chain — the actual error text is here instead of the generic "failed" toast.

Setup

First-run welcome

New installs start with a short three-step welcome and direct actions for opening a local video or using a web link. Editing stays on the computer, with no account or upload required.

Setup dialog

Opened from the ⚙ Setup button in the header. When FFmpeg is missing on first run, a setup screen explains the source, destination, and integrity check before asking permission to install. Each row describes a prerequisite and the feature it unlocks; required items are pre-checked, optional ones wait for opt-in. Select all missing toggles every installable row.

  • Install Selected runs in a background thread: FFmpeg is pulled as a portable Windows build from gyan.dev, checked against the publisher's SHA-256 digest, staged, and then placed in assets/ffmpeg/bin/ (the app's fallback lookup path). Python packages use python -m pip install --user; no admin access is needed.
  • Open Admin Terminal launches an elevated shell pre-populated with the right commands for your OS: winget install Gyan.FFmpeg on Windows (via PowerShell Start-Process -Verb RunAs), brew install ffmpeg on macOS (via Terminal + osascript), sudo apt-get install ffmpeg on Linux. If no terminal is available, commands are copied to the clipboard as a fallback.
  • Relaunch restarts the current process so newly-installed prerequisites are picked up.

Share (inside the Export dialog)

After a successful export, the Export dialog reveals a share panel with a caption entry and one button per supported platform. Clicking a button:

  1. Copies the exported file to the OS clipboard (Windows: PowerShell Set-Clipboard -Path; macOS: pbcopy; Linux: xclip)
  2. Opens the platform's compose / upload page in your browser (pre-filled with your caption on X, Reddit, and Facebook's sharer)
  3. Shows a one-line instruction ("Click + Create, then paste the file", etc.)

Text layers

Multiline text tools with the persistent feature dock

The editor exposes a collapsible Text Layers panel with per-layer controls:

  • Style presets — Subtitle (white-on-black box), Caption (white with black outline, the social-standard look), Title (large centered), Watermark (small corner), Custom
  • Per-layer font (all system fonts), size, color (8 presets + Custom… color picker), position (7 anchors)
  • Bold / italic toggles, resolved to real font-variant files (arialbd.ttf, ariali.ttf, ...) with graceful fallback when a variant is missing
  • Outline and Shadow toggles, compiled to drawtext borderw / shadowx and mirrored exactly in the preview
  • Multiline captions: the text box wraps, and embedded newlines export as real line breaks
  • Per-layer timing slider — exactly when each text appears and disappears
  • 🎯 Motion track — keyframe a caption's position so it follows a moving subject; drag to set the path or ⚡ Auto-track it with OpenCV, then compiled to a drawtext time expression that preview and export share
  • Background box toggle
  • ▲ / ▼ reorder, ⧉ duplicate, ✕ remove
  • Text fade (0.25s / 0.5s / 1s, set in Export Options) — symmetric fade-in/fade-out via a drawtext alpha= expression
  • Live PIL-rendered overlay on the preview canvas — font size, position, and box padding all match ffmpeg's export output

Quality presets

Preset FPS Max Width GIF Colors Video CRF
Low 10 480px 64 28
Medium 15 720px 128 23
High 24 1080px 256 18
Ultra 30 Native 256 15

When a hardware encoder is available, CRF maps to the right flag per encoder (-cq for NVENC, -global_quality for QSV, -q:v for VideoToolbox).


Keyboard shortcuts

Key Action
Space / K Play / Pause
J Seek −1s
L Seek +1s
I Set in-point at current frame
O Set out-point at current frame
Ctrl+Z Undo (text-layer edits, crop, trim range, queued ranges)
Ctrl+Y / Ctrl+Shift+Z Redo
Ctrl+E Export
Ctrl+O Open video file
Ctrl+V Paste — a video/GIF link opens the Kidnap downloader from any tab; a clipboard image becomes an overlay on Trim

Entry fields swallow shortcuts so typing into them doesn't scrub the video.


Installation

Option A — Windows (no Python required)

Get it from the Microsoft Store (signed, auto-updates, no security warning), or download VideoKidnapper.exe / the Setup installer from the latest release. If a prerequisite is missing, first-run setup explains the verified download and asks before installing it.

Option B — macOS (no Python required)

Download the .dmg for your Mac from the latest release…-macos-arm64.dmg (Apple Silicon) or …-macos-x86_64.dmg (Intel) — and drag the app to Applications. FFmpeg is bundled.

The app isn't code-signed yet, so on first launch right-click it → Open (a plain double-click is blocked by Gatekeeper for unsigned apps). After that it opens normally.

Option C — Linux AppImage (no Python required)

Download VideoKidnapper-x86_64.AppImage from the latest release:

chmod +x VideoKidnapper-x86_64.AppImage
./VideoKidnapper-x86_64.AppImage

FFmpeg is bundled — nothing else to install. Works on any glibc 2.35+ distro: Ubuntu 22.04+, Debian 12+, Fedora 36+, and immutable distros like Bazzite, SteamOS 3+, and Silverblue (where the AppImage is the recommended route since you can't layer packages). For an app-menu entry + icon, add it with Gear Lever.

On Ubuntu / Debian / Mint you can instead use the APT repository — one-time setup, then updates arrive through normal apt upgrade:

sudo install -d /etc/apt/keyrings
curl -fsSL https://aes256afro.github.io/apt/videokidnapper.asc | sudo tee /etc/apt/keyrings/videokidnapper.asc > /dev/null
echo "deb [signed-by=/etc/apt/keyrings/videokidnapper.asc] https://aes256afro.github.io/apt stable main" | sudo tee /etc/apt/sources.list.d/videokidnapper.list
sudo apt update && sudo apt install videokidnapper

(Or grab the .deb from the release page and sudo apt install ./videokidnapper_*.deb — same package, no repo setup, no auto-updates.)

PyPI also works if you prefer pip:

sudo apt install python3-pip python3-tk ffmpeg xclip
pip install "videokidnapper[all]"
videokidnapper

Option D — PyPI (recommended if you have Python)

pip install videokidnapper            # core install
pip install "videokidnapper[dnd]"     # + drag-and-drop support
videokidnapper                        # launches the GUI
videokidnapper --help                 # CLI mode

You still need FFmpeg on PATH (or use the in-app ⚙ Setup dialog to install a verified portable copy on Windows).

Option E — Clone and install (contributors / latest main)

1. Install Python 3.9 – 3.14

2. Clone and install

git clone https://github.com/AES256Afro/VideoKidnapper.git
cd VideoKidnapper
pip install -e .                      # editable install — picks up your edits
# or the old way:
pip install -r requirements.txt

3. FFmpeg

Three options — the Setup dialog handles all of them:

  • Auto-install (Windows): open ⚙ Setup → check FFmpeg → Install Selected. Pulls the gyan.dev essentials build into assets/ffmpeg/bin/.
  • Manual portable: drop ffmpeg.exe and ffprobe.exe into assets/ffmpeg/bin/ yourself.
  • System install: winget install Gyan.FFmpeg (Windows) / brew install ffmpeg (macOS) / sudo apt install ffmpeg (Linux).

4. Run

python main.py           # GUI (works from a clone without installing)
python main.py --help    # CLI help
# or, after `pip install -e .`:
videokidnapper           # GUI
videokidnapper --help    # CLI help

CLI mode

python main.py --url "https://youtu.be/..." --start 10 --end 25 \
               --format GIF --quality High --speed 1.5 --aspect 9:16

All flags: --url, --file, --start, --end, --out, --format {MP4,GIF}, --quality {Low,Medium,High,Ultra}, --speed, --rotate {0,90,180,270}, --mute, --audio-only, --aspect {Source,1:1,9:16,16:9,4:5,3:4}, --hw {auto,off}.

Passing any flag (or --help) skips the GUI and runs headless.


Export naming

Files are saved as VidKid_{mode}_{YYYYMMDD}_{HHMMSS}.{ext} in the configured output folder (set via Export Options → Output folder).

Example: VidKid_trim_20260417_221530.mp4

When "Concat queued ranges" is enabled, the final merged output uses _concat in the mode slot and the intermediate per-range files are cleaned up.


Tech stack

  • CustomTkinter — dark-themed GUI framework
  • Pillow — frame preview + live text-layer overlay
  • yt-dlp — multi-platform video downloading
  • FFmpeg — video/GIF encoding with drawtext overlays
  • mss — cross-platform screen capture
  • tkinterdnd2 (optional) — drag-and-drop on the preview canvas

Tests

pip install pytest
python -m pytest tests/ -v

500+ tests covering URL detection, project files and recovery paths, install-aware updates, verified prerequisite staging, platform share intents, ffmpeg filter construction, text styling, download retry classification and cookie resolution, settings migration, SRT parsing, size estimation, cache behavior, and drag-and-drop payload parsing.



Disclaimer & terms of use

VideoKidnapper is a personal utility for trimming and re-encoding video you have the right to use. By running this software you acknowledge:

  • Platform terms of service. Downloading from services like YouTube, Instagram, Twitter/X, Reddit, Bluesky, and Facebook may violate their terms of service. You are responsible for ensuring your use complies with each platform's current ToS, your local laws, and applicable copyright law (DMCA, Fair Use, and equivalents in your jurisdiction).
  • Copyright. Re-hosting, re-sharing, or monetizing content you don't own or have a license to use is your responsibility and not this project's.
  • Cookies. The "Cookies from browser" option reads authentication cookies from your installed browsers through yt-dlp. Never share a cookie file or export — it grants session-level access to your accounts.
  • No warranty. The software is provided "as is" without warranty of any kind; see the LICENSE file for the full terms.

The project's authors and contributors do not endorse or encourage violation of any platform's terms of service or any applicable law.


License

Licensed under the Apache License, Version 2.0 — see LICENSE for the full text.

Note on historic releases. The v1.0.0 tag and earlier were published under the GNU General Public License v3.0 and remain available under GPLv3. Apache-2.0 applies to v1.1.0 and every later commit on main.

VideoKidnapper depends on several third-party projects with their own licenses:

Project License
CustomTkinter MIT
Pillow MIT-CMU / HPND
yt-dlp Unlicense
mss MIT
tkinterdnd2 BSD-3-Clause (optional)
FFmpeg LGPLv2.1+ / GPLv2+ (build-dependent; essentials builds are GPL)

If you bundle FFmpeg with a redistribution of VideoKidnapper, note that the essentials and full gyan.dev builds ship under GPL — complying with GPL requires offering source on request. Alternatively, use an LGPL FFmpeg build for LGPL-only redistribution.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

videokidnapper-1.8.0.tar.gz (302.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

videokidnapper-1.8.0-py3-none-any.whl (281.6 kB view details)

Uploaded Python 3

File details

Details for the file videokidnapper-1.8.0.tar.gz.

File metadata

  • Download URL: videokidnapper-1.8.0.tar.gz
  • Upload date:
  • Size: 302.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.14

File hashes

Hashes for videokidnapper-1.8.0.tar.gz
Algorithm Hash digest
SHA256 94ec5622abb00fa8c53d6db9e2b99eb01f16f3b25c68738338dde048234679d3
MD5 f5bdee46cacf0f7be18683658fde7f2f
BLAKE2b-256 dedde18110fb3039401e9173b9c584da390a7dfe4676ebf689130a9a5486434f

See more details on using hashes here.

Provenance

The following attestation bundles were made for videokidnapper-1.8.0.tar.gz:

Publisher: release.yml on AES256Afro/VideoKidnapper

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file videokidnapper-1.8.0-py3-none-any.whl.

File metadata

  • Download URL: videokidnapper-1.8.0-py3-none-any.whl
  • Upload date:
  • Size: 281.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.14

File hashes

Hashes for videokidnapper-1.8.0-py3-none-any.whl
Algorithm Hash digest
SHA256 aa903946cb0da81523be9f2e3bc4cab7025665628d7334859f28f80616ef746b
MD5 6c98557f532fb71d94b7973406c3b880
BLAKE2b-256 df7353822379a719b62814a17ec45386ecf2c09f9b804854b95cc25569fde001

See more details on using hashes here.

Provenance

The following attestation bundles were made for videokidnapper-1.8.0-py3-none-any.whl:

Publisher: release.yml on AES256Afro/VideoKidnapper

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page