Transcribe Video to Text Online
Pull spoken words out of video files into searchable text, captions, or structured JSON.
Whether you are adding accessibility captions, building a searchable archive, or just want to skim a long video, transcription unlocks the speech inside your file. Media Manipulator transcribes your video locally on our own GPU server — no third-party AI provider sees your media.
How to use it
Upload video
Drop in any MP4, MOV, WebM, or MKV file.
Extract audio
FFmpeg pulls a clean audio track from the video.
Local AI transcription
whisper-ctranslate2 runs on our GPU server.
Download transcript
Save as VTT captions, plain text, or JSON.
What this tool does
- Extracts audio from your video and runs speech-to-text on it.
- Exports captions in WebVTT (.vtt), plain text, or structured JSON.
- Supports many spoken languages with automatic language detection.
- Runs locally on our GPU server — no third-party AI provider involved.
Why it matters
- Captions make your video accessible to deaf and hard-of-hearing viewers — and to anyone watching on mute.
- Search engines and AI agents can index a transcribed video far better than a raw video file.
- Internal teams can search across recorded calls, demos, and tutorials when transcripts exist.
Common use cases
Adding captions
Generate VTT files for accessibility and silent autoplay.
Searchable archive
Index recorded meetings, calls, or training videos.
Content repurposing
Turn a video into a blog post or summary with a quick start.
Compliance and notes
Keep written records of recorded interviews or interactions.
Advanced details
- Transcripts come from whisper-ctranslate2 (a faster Whisper implementation) running on a local GPU we operate.
- VTT output is the right pick for video captions — it works with HTML <track> elements, YouTube, and most video editors.
- JSON output includes per-segment start/end timestamps and confidence values, useful for building search indexes or summarization pipelines.
Why Media Manipulator
- Runs entirely on our own GPU server — no OpenAI/Anthropic/Google sees your audio.
- Outputs the format you actually need: VTT, text, or JSON.
- Same upload + delete-within-24-hours privacy model as the rest of the toolset.
Privacy & file handling
Your uploads are processed on our own servers and are designed to be automatically deleted within 24 hours. We do not share your files with third-party AI providers, and AI features run on a local GPU server we operate. No login or account is required.
Frequently asked questions
Which video formats are supported?
MP4, MOV, WebM, MKV, AVI, FLV, and most other common containers. Audio files can also be transcribed directly.
How accurate is the transcript?
Word-error rates are typically a few percent for clear speech. Heavy background music, overlapping voices, or strong accents lower accuracy.
Can I get timestamps?
Yes — VTT and JSON outputs both include per-segment timestamps.
Does my video leave your servers?
No — transcription happens on a GPU server we operate. The original file and the transcript are deleted within 24 hours.