Subtitle Word Count
Processed locally in your browser — nothing is uploaded.
Quick answer
Subtitle Word Count reads a caption file and reports its statistics: total words, characters and cues, the file’s total duration, the average reading speed in characters per second, and the longest line. It runs entirely in your browser — the file is parsed in memory and never uploaded — so you get an instant, accurate count for translation quotes, dubbing estimates and quality checks.
How to subtitle word count
Load the subtitle file
Drag any .srt, .vtt, .ass, .sbv or .ttml file in, or paste the text. The format is detected automatically.
Read the statistics
The tool counts words, characters and cues, sums the on-screen time, and calculates the average reading speed and the longest line.
Use or copy the numbers
Copy the figures for a quote or a QC note. Nothing is stored or sent — the counts are computed on your device.
What does it report?
The tool parses the file into cues and measures both the text and the timing, so you get a rounded picture rather than a single number:
| Metric | What it means |
|---|---|
| Words | Total spoken words across all cues, tags excluded |
| Characters | Total characters, with and without spaces |
| Cues | Number of individual caption blocks |
| Total duration | From the first cue’s start to the last cue’s end |
| Average reading speed | Mean characters per second across the file |
| Longest line | The single line with the most characters |
Why word count matters for translation quotes
Subtitle translation and localisation are almost always priced per word or per subtitle, so the word and cue counts are the raw inputs to a quote. Before you send a file to a translator — or accept a bid — knowing it is 4,200 words across 780 cues tells you immediately whether a quote is fair. It also lets you compare vendors on the same basis, since both are quoting against the same measured file rather than a guess.
Estimating dubbing and voice-over length
The total duration and average reading speed are what you need for audio work. Duration tells a voice actor or dubbing studio how much timed content there is, which drives session length and cost. The average CPS is a sanity check on pacing: a very high figure warns that the script is dense and may be hard to voice — or to read — at the current timing, and a very low figure suggests cues that linger.
Using the numbers for quality control
The longest line and average reading speed double as QC flags. Broadcast guidelines put a comfortable line at around 42 characters and a comfortable speed at roughly 17 characters per second; a longest line well past that, or an average CPS that is high, points to captions that will be hard to read. To find and fix the specific offending cues rather than just the averages, open the file in the subtitle timing checker, which flags every cue individually.
Is the file uploaded to count it?
No. The counting is arithmetic performed in your browser — the file is parsed and measured in memory, and nothing is transmitted. Check the Network tab in DevTools and you will see no upload while the statistics are computed. Confidential scripts and unreleased captions stay entirely on your device.
Related subtitle tools
Need the words as a document rather than a count? Use subtitle to text. To act on a high reading speed, retime cues in the subtitle editor. And if you are quoting a dub, you may first want to extract the audio from the video to hand to your voice talent.
Your files never leave your device
Processed locally in your browser — nothing is uploaded. You can verify it yourself: open your browser’s developer tools, watch the Network tab, and use the tool — no upload request is ever made.
Frequently asked questions
How are words counted — including the timecodes?
No. Only the spoken caption text is counted; cue numbers, timecodes and styling tags are excluded, so the figure reflects the words a translator or reader actually deals with.
What counts as a cue?
Each individual caption block — one timing range and its text — is one cue. The cue count is what per-subtitle pricing is based on.
How is the average reading speed calculated?
It divides the total characters shown by the total on-screen time to give a mean characters-per-second figure across the whole file, a quick gauge of pacing.
Does it count characters with or without spaces?
Both are reported. Some rate cards use characters with spaces and some without, so you have the figure either way.
Can I use this to quote a translation job?
Yes — that is a core use. Multiply the word count by a per-word rate, or the cue count by a per-subtitle rate, for an instant, defensible budget.
Why is my longest line flagged as a problem?
It is not automatically wrong, but lines much longer than about 42 characters wrap awkwardly on small screens. The figure is there so you can spot outliers worth splitting.
Does it work on ASS and VTT, not just SRT?
Yes. SRT, VTT, ASS/SSA, SBV and TTML are all parsed into the same model before counting, so the statistics are consistent across formats.
Is my file uploaded to be measured?
No. All counting happens locally in your browser. You can confirm there is no network request in your developer tools.