Transform Long Videos into Clips in One Click
Say goodbye to monthly subscriptions. A unified workspace brings import, transcription, clip studio, and export together, with word-level snapping and 16-language transcription. Process your data locally with complete privacy or in the cloud at minimal cost.
Windows
Linux
macOS (Coming Soon)
Who is Clippat For?
Content Creators
Turn streams and podcasts into Reels and Shorts quickly.
Video Editors
Ready-to-use SRT files and Rough Cuts that save hours of work.
Students
Transcribe lectures into text ready for summarization and study.
Journalists
Transcribe interviews accurately and with 100% offline secrecy.
Core Features
Unified Projects Workspace
Import a video, read the full transcript, and shape clips in one studio, then export from the same screen.
Waveform & Word-Snap Studio
Set clip start and end on the waveform at word precision, with silence gaps and live previews of both edges.
16-Language Transcription with Diarization
Transcription engines cover Arabic dialects (Levantine, Egyptian, Gulf, Maghrebi), English, and 14 more languages, with speaker labels.
Compilation Strip
Reorder clips with drag and drop, preview them in sequence, and export one merged video.
Series Recording Sessions
Merge multiple recordings into one session in a custom order, with consistent loudness normalization.
Bundled Runtimes
Clippat 2.0 ships with FFmpeg, Python, and llama-server, so transcription and analysis run locally without manual setup.
Workflow
From raw video file to engaging clips in 4 steps
Select File
Supports various video formats and MP3 audio files.
Speech Transcription
Choose local GPU processing or cloud processing via Gemini API.
Text Processing
AI processes transcript text and suggests optimal clip cuts.
Clipping & Export
Get high quality video clips complete with suggested titles and timestamps.
Full Support for Arabic Dialects
Gulf
Local term support
Egyptian
Colloquial expressions
Levantine
Mixed Arabic/English
Maghrebi
Mixed French support
Ownership vs Rental
The core difference between Clippat and other services is the business model. We sell you software; they sell you time.
Lifetime License
Pay once, use forever.
100% Privacy
Offline local processing for your sensitive data.
No Minute Limits
Process hundreds of hours locally with no extra fees from us.
| Feature | SaaS | Clippat |
|---|---|---|
| Payment Model | Monthly | One-time |
| Privacy | Cloud | Local |
| Internet | Required | Periodic license check only |
| Dialects | Not supported | Full Support |
Your Data Never Leaves Your Device
In "Local Processing" mode, Clippat runs in a fully isolated environment: your sensitive audio files, interviews, and exclusive content are processed entirely on your device hardware, with a simple setup wizard that lets you choose Whisper or Cohere models for transcription and GGUF models for analysis.
100% Local Privacy
Guided Setup • AI Models
User Reviews
"Clippat is a promising app that stands out with its high transcription accuracy across different dialects and powerful clip creation while maintaining topic cohesion. The best part is that it allows seamless Google cloud integration if your device hardware is limited."
Twitter User
Transcribing religious lectures in Maghrebi dialect
Technologies & AI Models
Built on local and cloud models for Arabic and English transcription and clip analysis, with bundled runtimes ready to work
Single-Click Model Installation Wizard
Clippat 2.0 ships with FFmpeg, Python, and llama-server, downloads the models suited to your hardware, and configures them without complex setup.
Faster-Whisper & CTranslate2
High-performance local speech transcription with low memory footprint and hardware acceleration.
Google Gemini API (2.0 / 2.5 / 3.1)
Lightweight cloud integration providing exceptional transcription accuracy even on lower-end devices.
Local LLMs (llama.cpp / GGUF)
Run local AI models through llama.cpp to analyze transcripts and extract video clips with 100% privacy, with OpenRouter support for cloud models.
Hardware & OS Support Matrix
Comprehensive compatibility across desktop devices and operating systems
| Hardware / Platform | Operating Systems | Acceleration Engine | Recommended Specs | Status |
|---|---|---|---|---|
| NVIDIA GPUs | Windows / Linux | CUDA / TensorRT | 4GB+ VRAM (RTX / GTX 1060+) | Full Optimal Support |
| AMD GPUs | Linux / Windows | ROCm / DirectML | 6GB+ VRAM (RX 6000+) | Supported |
| CPU / Cloud Fallback | Windows / Linux | AVX2 / Gemini API | Any CPU + Light Internet Connection | Universal Support |
Frequently Asked Questions
Direct answers about AI clipping, speech transcription, offline operation, and licensing.
What is the best AI video clipping app for Arabic content?
Clippat is a desktop app for AI video clipping that runs on Windows, Linux, and macOS. It understands Modern Standard Arabic and the Gulf, Egyptian, Levantine, and Maghrebi dialects, transcribes the audio, suggests publishable clips, and exports them in a 9:16 vertical frame with an SRT subtitle file. It runs in a local mode that uploads no file to any server.
How do I turn podcasts and long videos into Reels, Shorts, and TikTok clips?
Import the video or audio file into the workspace, then choose the transcription language and the Whisper model that fits your hardware. Clippat produces a word-level timed transcript and shows it in the clip studio next to the waveform. Set the clip start and end, adjust the 9:16 vertical frame, then export an MP4 with burned-in or sidecar subtitles. The same steps cover Reels, Shorts, and TikTok without changing any settings.
Is Clippat an OpusClip alternative that works offline?
Yes. In local processing mode, transcription and scene analysis run on your machine through Faster-Whisper and GGUF models with llama.cpp, so no internet connection is needed once the models are downloaded. A connection is still required to verify the license and for optional cloud processing through Gemini or OpenRouter.
Which languages and transcription models does version 2.0 support?
Version 2.0 transcribes speech in 16 languages, with dedicated handling for Arabic and its Gulf, Egyptian, Levantine, and Maghrebi dialects, plus speaker diarization for recordings with multiple voices. Available Whisper models are Tiny, Base, Small, Medium, Large-v3, and Turbo, accelerated by Faster-Whisper and CTranslate2, alongside cloud Gemini models for low-spec hardware.
What are the hardware requirements for running local models?
Clippat runs on Windows 10 or later, Ubuntu 22.04 or later, and macOS 13 or later. CPU processing with 8 GB of RAM handles the smaller models, while 16 GB is preferable for Large-v3. An NVIDIA GPU with 4 GB of VRAM or more (GTX 1060 or newer) accelerates transcription through CUDA, and an AMD GPU with 6 GB of VRAM or more (RX 6000 or newer) runs through ROCm or DirectML.
What is the difference between the free trial and the full license?
The trial gives you seven days of full features, starts at your first sign-in, and needs no credit card. The full license is a one-time payment of 30 USD, does not renew, carries no monthly fees, and keeps working after the trial period ends.
Start Your Free Trial
No credit card required. Your 7-day free trial starts when you sign in.