Video Subtitles – automatically generated in SRT, VTT and YouTube format
Upload MP4, MOV, MKV – finished subtitles in 99+ languages, from €0.18/hour of video.
3 free transcriptions · no credit card · data stays in Germany
SRT · VTT · TXT · JSON
Subtitles double a video's reach: muted consumption on LinkedIn/Instagram, accessibility, SEO indexing of spoken content. DeepScript accepts MP4, MOV, MKV, WebM directly and delivers SRT + VTT with word-accurate timestamps – upload to YouTube or import into your video editor.
Recommended setup
How to get the best result
For this use case we recommend the Standard model – here are the concrete settings.
- Model
- Standard · €0.18/hour
- Language
- Set the speakers' language, or auto-detect for multilingual videos.
- Export
- SRT (universal standard, every NLE reads it), VTT (HTML5/YouTube native), JSON when you need word-level highlighting in your player.
Cost example
What this costs in practice
Example: 12-minute LinkedIn explainer video, EN, Standard model. Cost: €0.04 (minimum charge €0.01 – only 12 min here). Processing time: ~45 sec. Output: 80 subtitle cues in SRT + VTT, word-accurate timing.
How it works
From recording to transcript in a few steps
- 1
Upload the video
MP4, MOV, MKV, WebM up to 5 GB. We extract the audio track automatically; the video itself isn't modified or stored.
- 2
Set the language or auto-detect
Multilingual videos: auto-detect runs per-segment and produces a mixed transcript.
- 3
Download SRT/VTT
Upload directly to YouTube (subtitles tab) or import into your NLE (Adobe Premiere, DaVinci Resolve, Final Cut).
FAQ
Frequently asked questions
Which video formats are supported?+
MP4, MOV, MKV, WebM, AVI. We extract the audio track via ffmpeg – anything ffmpeg can read works.
Can I configure subtitle cue length?+
Default is 2-7 seconds per cue (standard reading speed). For tighter or wider limits take the JSON output and re-segment with your own tool.
Can I auto-translate the subtitles?+
We no longer offer our own translation API – instead we expose an MCP endpoint so Claude or ChatGPT can translate directly on the transcript (significantly better than generic MT services). Alternative: DeepL/Google after export.
How does background noise affect the result?+
Premium model has better denoising. For very loud background (music, crowd noise) we recommend a quick preprocess in Audacity or iZotope RX before upload – meaningful accuracy boost.
Try it now?
Upload a file and see the result in minutes. Three transcriptions free, no credit card, data stays in Germany.