Subtitle creation
Convert video to an editable SRT file
Generate timed subtitles for DaVinci Resolve, YouTube, social clips, courses and accessible video. The video stays on your device while Whisper creates an editable word-level transcript.
Quick answer
Whisper Online downloads a whisper.cpp model, processes the selected media locally, lets you correct individual timed words, and exports the result without uploading the media file.
Step by step
The practical workflow
- 1
Select a local video
Choose your MP4, MOV, WebM or another browser-readable video. Audio is extracted locally rather than uploaded to a transcription server.
- 2
Run the Whisper model
Pick a model based on download size, speed and accuracy, then let the browser produce individual timed words.
- 3
Edit text and timing
Correct the paragraph directly or use Detailed edit for a row-by-row list of tokens, start times and end times.
- 4
Shape and download captions
Export SRT with limits for words, characters, lines and duration. Import it into an editor or upload it with the video.
One subtitle source for several destinations
SRT is widely used because it keeps subtitle text and timing in a small, portable file. A single reviewed transcript can become captions for an editing timeline, a hosted video, a course lesson or an offline media player.
Whisper Online preserves individual word timing while you edit. The SRT exporter then groups those words into readable caption cues according to your chosen limits.
- DaVinci Resolve and other nonlinear video editors
- YouTube and other platforms that accept caption files
- Tutorials, product demos and online courses
- Interviews, documentaries and talking-head videos
- Accessibility captions for recorded events
Choose caption limits for the screen
Subtitle readability depends on the destination. Horizontal video has more line width, while vertical video competes with buttons, titles and safe-area margins. Export settings let you create separate files without retranscribing.
Use character limits to control line width, word limits to keep captions conversational, and duration limits to prevent a cue from staying on screen too long. Always watch the result against the final edit before publishing.
Private video transcription without an upload queue
Many online subtitle generators send the complete media file to a remote service. Local transcription is different: the model and inference code run in the browser, so drafts, client footage and interviews remain on the device.
Performance depends on the computer, browser, file duration and selected model. Tiny is the lightest starting point; larger models require more memory and processing time.
Why use local transcription?
FAQ
Common questions
Can I convert MP4 to SRT?
Yes. Select the MP4, transcribe its audio locally, review the timed words and choose SRT from the export controls.
Can I edit subtitle timings before downloading?
Yes. Detailed edit presents every token in a list with editable text, start time and end time. You can also add and remove tokens.
What is the difference between SRT and VTT?
Both contain timed captions. SRT is commonly used by video editors and players, while WebVTT is designed for web video and supports browser-oriented cue syntax.
Is there a file upload limit?
The app does not impose a server upload limit because media is not uploaded. Practical limits come from browser memory, decoding support and the performance of the device.
Related guides