Version 2.0.0 • One-time download • Runs offline • Windows
Generate precise, word-level .SRT subtitles — fast.
Polley Word by Word SRT is a desktop app that converts audio or video into word-level subtitle files you can use in Premiere Pro, DaVinci Resolve, CapCut, Shorts, Reels, and TikTok.
Drop in one file or a whole batch, choose an accuracy, and get timed words you can correct word by word — then export one word per subtitle straight to .SRT, or burn styled captions into an MP4.
This is a one-time download, runs entirely offline, and stores output locally on your PC. No Python. No separate FFmpeg install. No developer tools.
Input
Drop one clip or a whole batch of audio and video.
Output
Word-level .SRT files, or a captioned MP4, written wherever you choose.
The app corrects each word, lets you re-time it on a timeline, and previews the result on the actual video before you export.
Demo video
Quick walkthrough of Polley Word by Word SRT.
Runs locally on Windows. Generates word-level .srt files you can import into your editor.
Get on itch.ioWhat it does
- • Converts audio/video into a word-level .srt subtitle file
-
• A full caption editor, not just a converter:
- ◦ A synchronized video and audio preview
- ◦ Word-by-word correction, undo and redo
- ◦ A word timeline for exact start and end times
- ◦ Search and replace across every caption
- • Batch process multiple files and resume if interrupted
- • Burn styled captions into an MP4 with GPU acceleration
- • Save projects and recover unfinished work after a crash
Best for
What this is not
- • Not a cloud service
- • Not a subscription
- • Not an online upload tool
- • Not something that needs Python or FFmpeg
It’s a local app built for speed and clean output. It installs and runs on a normal Windows PC.
How it works
One clip or a whole batch, corrected word by word.
Steps
- 1. Drop in one audio/video file, or a New Batch of several
- 2. Pick an accuracy — Fast, Balanced, or Accurate — and press Start analysis
- 3. Play the preview and fix any word the model got wrong — edit, delete, add, or search-and-replace
- 4. Nudge timing on the word timeline if you want frame accuracy
- 5. Export word-level .SRT — one word per subtitle — or a captioned MP4
What you can change
- • Words and their timing
- • Grouping: one word, phrase, or sentence
- • Font, size, colors, outline, and background
- • Placement, plus Shorts/TikTok/Reels safe guides
- • Entrance and exit transitions and fades
- • Word highlight or enlarge while a word is spoken
Where output goes
You pick the folder. Batch exports write one UTF-8 .SRT per reviewed item, and captioned videos render into the folder you choose — nothing is trapped in a fixed uploads folder.
Notes
- • The first analysis downloads a verified speech model
- • Every analysis after that runs locally
- • CPU usage is normal while processing
- • Closing mid-batch is safe — it can be resumed
Use cases
Where word-level subtitles shine.
Short-form
Snappy captions with tight timing for Shorts, Reels, and TikTok.
Karaoke
Word-by-word highlight effects that follow speech precisely.
Editing
Import .srt into your editor and build caption styles fast.
FAQ
Everything customers ask before buying, answered plainly.
How do I generate word-by-word subtitles from audio or video?
Open Polley Word by Word SRT, drop in a file, choose an accuracy, press Start analysis, correct any words you need to, then export the .SRT. Word-level is the default, so the file comes out with one word per subtitle.
Does it upload my audio/video to the internet?
No. It runs locally on your Windows PC and your files stay on your computer. The only request it ever makes on its own is a check for a newer version.
Is it a subscription?
No. It’s a one-time download (not a subscription or cloud app).
Where are my files saved?
Wherever you choose. Exports ask for a folder and write your .SRT files or captioned videos there. The app’s own settings, presets, and downloaded models live in your Windows user folder, not alongside your media.
What if I select multiple files?
Use New Batch. Each clip keeps its own transcription and corrections while sharing your caption style. You review them one by one, then export every ready item into a folder you pick. If you close mid-batch, it picks up where it left off.
What are the system requirements?
A 64-bit Windows 10 or Windows 11 PC. That’s it — no Python, no separate FFmpeg installation, and no developer tools. The speech engine and media tools ship inside the app. The first analysis downloads a verified speech model (about 32 MB for Fast, 60 MB for Balanced, 190 MB for Accurate); after that everything runs offline.
Can I just burn the captions into the video instead?
Yes. Export a captioned MP4 with your chosen style, placement, fades, and transitions. Hardware encoding is used automatically when your graphics card supports it, with a CPU fallback.
I already have subtitles — can I use those instead?
Yes. Import an .SRT or .VTT file and its cues become timed words you can correct, regroup, restyle, and export like any other project. This also works on silent video.
Ready to generate word-level subtitles?
Download once. Runs offline.