SRT, VTT, ASS · one editor

Subtitles to speech.
Whichever file you already have.

SRT, WebVTT, and ASS all land in the same editor. The cue times become the speech times, and a name carried by the format — a WebVTT voice tag or an ASS Name — can keep its own voice.

10 free minutes · No credit card required

How any of the three becomes speech

  1. 1

    Upload the subtitle file you have

    SRT, VTT, and ASS are accepted directly. There is no convert-to-SRT step and no timeline to rebuild.

  2. 2

    Check speakers and wording

    Names the format already stored show up as speakers. Delete on-screen text you do not want read aloud.

  3. 3

    Generate into the cue windows

    Pick a voice per speaker, in any of 57 languages after a translation. Download the speech, or the video with that audio.

What each format brings with it

The parsers differ. The speech step does not: every format becomes cues with a start, an end, and a speaker. Format-specific notes live on the SRT, VTT, and ASS pages.

SRT
A cue number, a comma timestamp, and text. Most SRT files are one speaker. A Zoom-style Name: prefix is promoted to a speaker only when the pattern is consistent: at least two cues, at least half of them named, and at most 12 distinct names.
WebVTT
A <v Name> tag becomes the speaker. Several voice tags in one cue become separate lines, and that cue’s time is divided by how many characters each voice says. Header, notes, style, regions, and positioning are dropped because they describe a screen, not speech.
ASS and SSA
The Name column on a Dialogue line is the speaker. Style names are layout, so they are not treated as people. Override tags and karaoke timing are stripped and the words inside them are kept. A .ssa file is the same layout; rename it to .ass to upload it.
Cues with no name
Lines that name nobody share one speaker of their own, so you can see them separately from the named people. A file that names nobody at all stays a single speaker.
No subtitle file yet
Upload the audio or video instead. Transcription builds the cues, up to 30 minutes, and those cues are what the speech is timed to. Transcription itself is not billed against voiceover minutes.
Same limits after import
Whichever format you upload, the subtitle file is read up to 1 MB, and the generated timeline can run up to 30 minutes.

Pick the file you were given

An SRT from a subtitle editor

SubRip is the file most editors and caption tools still hand over. Upload it when the timing is already approved and you want it spoken.

How an SRT cue is spoken

A WebVTT export with speaker tags

YouTube, meeting tools, and HTML5 players often deliver WebVTT. Voice tags arrive as separate speakers, ready for separate voices.

What a VTT file keeps

An ASS script from Aegisub

Fansub and typesetting scripts store the speaker in the Name column and the styling elsewhere. The name is kept; the styling is not read aloud.

What an ASS file keeps

Only the video, no captions

Skip the hunt for a subtitle file. Transcribe the recording, edit the lines, and generate speech into the times the transcript just produced.

Start from the recording

One speech step for three subtitle formats

You do not pick a different product for SRT, VTT, and ASS. Each parser keeps the parts that matter for speech — times, words, and any speaker name the format actually stores — and drops the parts that only place text on a screen.

Speaker names ride along when the format has them

WebVTT voice tags and the ASS Name column become speakers you can assign. SRT usually has no name column, so those files stay one speaker unless the cues themselves carry a consistent Name: prefix.

Translate after the format is parsed

The 57 translation languages apply after import, whichever subtitle format you started from. The cue windows stay; the words change; the speech is fitted to those windows again.

Watch it work

Frequently asked questions

Which subtitle formats can I upload?
SRT, WebVTT (.vtt), and Advanced SubStation Alpha (.ass). A classic .ssa script uses the same event layout; rename the extension to .ass and upload it. The file is read up to 1 MB.
Do I have to convert VTT or ASS to SRT first?
No. Each format is parsed on its own. Converting first would throw away speaker tags and, for ASS, the Name column, which is the information the voices are assigned from.
How are speakers chosen when the file names them?
WebVTT uses the name inside a voice tag. ASS uses the Name column, never the Style. SRT and VTT cues that start with a Zoom-style Name: prefix are promoted when at least half of at least two cues carry one and there are no more than 12 names. Everyone else shares a single unnamed speaker.
What if I do not have subtitles?
Upload the audio or video. Transcription produces the timed cues, up to 30 minutes, and speech is generated from those. Transcription is not charged against your voiceover minutes. The recording workflow is on the replace-voice page.
Where do I read about the MP3 itself?
The SRT-to-MP3 page covers the file you download: one timeline, the encode, and the other containers in the format menu. The speech is the same whatever the extension.

Speak the subtitles you have

10 free minutes included · No credit card required