Audio & media · focused workspace

Subtitle Speaker Labeler

Assign speaker names by cue number, generate a speaker-labeled SRT and voice-tagged WebVTT, and export a transcript grouped by speaker.

Speaker assignmentsEditable projectsLocal processing

Source → inspect → export

Open project

Preparing the worked example…

Your source & settings

Settings & options

Open your media file

Files stay in your browser. Try the built-in sample first.

Choose file

Source text is limited to 1.5 million characters per editor; imported text files are limited to 1.5 MB.

Inspect your result

PREVIEW & EXPORT

Output file

Know what changed

Manual cue assignments, not automatic voice recognition. WebVTT text and names are escaped. Original subtitle styling tags become literal text in voice-tagged output.

Choose another output above or open Export files.

Explore the example

Assign speaker names by cue number, generate a speaker-labeled SRT and voice-tagged WebVTT, and export a transcript grouped by speaker.

What you can inspect

Speaker assignments — Cue, Speaker, Start, Text.

Keep the result you need

Use the output selector to inspect each generated file, then download it. Save a project to keep source inputs and settings together. Result files and editable projects serve different purposes.

Processing & file support

Source text is limited to 1.5 million characters per editor; imported text files are limited to 1.5 MB.

Your working session

Processing runs locally in a worker with a 30-second cancellation limit. Undo and redo keep up to 20 edits in this tab. Project files include imported assets and are limited to 40 MiB. There is no automatic saving.

Public pages include advertising and optional analytics under our privacy policy. Review the method and any flagged output before sharing files.

Using Subtitle Speaker Labeler

Assign speaker names by cue number, generate a speaker-labeled SRT and voice-tagged WebVTT, and export a transcript grouped by speaker.

Start with the worked example in srt subtitles and cue-to-speaker json, then paste your own source or open a text file.
Adjust unassigned speaker label. Change one setting at a time so you can understand its effect.
Choose “Process source”. Inspect the speaker assignments before using the result.
Download the selected output, inspect other files in Export files and save a project if you want to reopen the same inputs.

Example

Try the built-in example, inspect speaker assignments and compare the result after changing one option.

Questions & answers

What does Subtitle Speaker Labeler do?

Assign speaker names by cue number, generate a speaker-labeled SRT and voice-tagged WebVTT, and export a transcript grouped by speaker.

What are the important limitations?

Manual cue assignments, not automatic voice recognition. WebVTT text and names are escaped. Original subtitle styling tags become literal text in voice-tagged output.

Can I save and reopen my work?

Yes. Save a project JSON file containing your inputs, settings and any imported assets. Projects are limited to 40 MiB. Undo/redo stores up to 20 edits in this session and is not saved in the project. There is no automatic saving.

How do imports and output previews work?

Source text is limited to 1.5 million characters per editor; imported text files are limited to 1.5 MB. Text previews show up to 150,000 characters; downloads include the complete output. Tables show 50 rows per page; CSV files include all rows.