Log inStart for free
AI subtitle generator

Add localized subtitles automatically

Pick a track with speech and AI writes subtitles viewers can follow on mute, as spoken or in any of 94 languages. Every line appears the moment its first word is said.

Add subtitles to a videoSee how it works
Free to start · No credit card required
Before and after

Stop timing captions by hand

Captioning used to mean typing, timing and splitting every line. Here the timing comes from the words themselves.

Captioning by hand

  • Type every line while you scrub back and forth
  • Nudge each caption until it meets the speech
  • Split run-on lines by hand so they fit the frame
  • Send the lines to a translator, then retime them

In Vidwip

  • The AI agent transcribes the speech on the track
  • Every word gets its own start and end time, so lines start on cue
  • Lines break at pauses and speaker changes; as-spoken lines fit in 42 characters
  • Pick one of 94 languages; each line keeps its original timing
How it works

From a track to timed subtitles

Make subtitles on the track you are cutting. No caption app, no export, no re-import.

01

Pick the track

Open Subtitles, press Generate with AI and choose the Track. The flyout picks one with sound for you; switch it if the speech is elsewhere.

02

Choose a language

Keep it As spoken, or pick one of 94 languages. The AI agent translates line for line on the original timing.

03

Generate and restyle

Press Generate. Lines land on a subtitle layer above your video: retype any one, restyle the set, then export them burned in.

Features

Subtitles that sit on the word, not near it

Word-level timing, line rules built for reading, and every caption editable right on your timeline.

Word-level timing

Subtitles in sync on the first pass

Every word gets its own start and end time, so each line appears the moment its first word is spoken. You skip the slow part of captioning: nudging every line to meet the speech.

  • The spoken language is detected for you
  • Up to 30 minutes of speech per transcript
  • Lines land on a subtitle layer above your video
Subtitle translation

Translate without retiming a line

The subtitle translator puts lines made from speech into one of 94 languages, one for one, on the same timing. You ship a localized cut without touching a timestamp; uploaded .srt and .vtt files can't be translated.

  • One line in, one line out
  • Start and end times stay put
  • Burned into the export at up to 1080p
Reading rules

Lines viewers read at a glance

A new line starts at a sentence end, a speaker change or a 0.8 s pause, and as-spoken lines stay within 42 characters and 6 seconds on screen. Viewers keep up without rereading.

Word highlight

Every word lights up as it's said

Choose Create subtitles on an audio block, and 8 of its 13 looks, like Bold Yellow, highlight each word as it's spoken. Your captions hold attention the way short-form viewers expect.

Inline editing

Fix one line, or restyle them all

Click a caption to retype it or change its From and To times, or select every caption on the layer and change the style once. A misheard name takes one click to fix.

SRT and VTT import

Use the subtitles you already have

Upload a SubRip or WebVTT file up to 5 MB and it lands at its own timings, restyles like any caption and burns into the export. A translator's file gets your look without retyping a line.

The models

Two models, one set of subtitles

ElevenLabs Scribe v2 hears and times every word, up to 30 minutes per transcript. Pick a target language and DeepSeek V4 Flash translates the lines; As spoken skips it.

  1. 01

    ElevenLabs Scribe v2

    ElevenLabs

    Transcribes the track, detects the spoken language, times every word and labels speakers so lines break between them.

    • Word timestamps
    • Speaker labels
    • Auto language
  2. 02

    DeepSeek V4 Flash

    DeepSeek

    Translates each subtitle line one for one into the language you pick. The original start and end times stay put.

    • Line for line
    • Keeps timing

Subtitle the same clip again in this tab and Scribe's words are reused, not transcribed twice.

Who it's for

Built for speech that has to be read

Talking heads, interviews, lessons and ads: if someone speaks on the track, it can carry subtitles.

Short-form creators

Subtitle a 9:16 talking-head clip, move the lines to Middle in a bold font, and it still lands with the sound off.

Two-person interviews

A new line starts whenever the speaker changes, so host and guest never share a caption.

Course creators

Subtitle a lesson in stretches of up to 30 minutes and fix any misheard term in the Caption box.

Marketing teams

Make a Spanish and a Japanese cut of one ad: generate one language, export at up to 1080p, then run the next.

Localization teams

Upload the translator's .srt or .vtt, restyle it to match the video and burn it into the export.

Product marketers

Record your screen with your mic in the editor, then subtitle the narration without leaving the tab.

FAQ

Questions about AI subtitles

Files, languages, accuracy, editing and export, including the SRT question.

What is an AI subtitle generator?
An AI subtitle generator turns the speech in a video into timed on-screen text, often called AI captions. In Vidwip, ElevenLabs Scribe v2 transcribes the track you pick, times every word and labels speakers, then adds subtitle lines to your timeline. Keep them as spoken or translate them, edit any line, and export the video with them burned in.
How do I add subtitles to a video?
Put the video or audio on the timeline, open Subtitles and press Generate with AI. Pick the Track and a Language, then press Generate, and the auto subtitle generator adds the lines in the plain Caption look. For the 8 word-highlight looks, select an audio block and choose Create subtitles, which offers all 13 looks. A video uploaded from your device has no separate audio block, so word-highlight looks need a separate audio file of the same take on an audio track.
What files and lengths can it subtitle?
MP4, MOV or WebM video, or MP3, M4A, WAV, OGG, FLAC or WebM audio, up to 2 GB a file, including a recording made in the editor. Each transcript covers up to 30 minutes of a clip, starting anywhere in the first 4 hours of its file. Split or trim a longer take into stretches of 30 minutes or less. Links can't be used; add the file itself.
Which languages does it support?
Scribe v2 detects the spoken language on its own, so there's nothing to set. For translated subtitles, pick one of 94 languages, from Spanish, Hindi and Japanese to Swahili, Yoruba and Zulu. DeepSeek V4 Flash translates each line one for one and keeps its timing. You can't choose the source language or add a glossary, and each run makes one language.
What if it mishears a word or a name?
Click the caption and retype it in the Caption box; the preview updates as you type. You can't add a glossary to teach it names or terms in advance, so check proper nouns and product names before you export. Scribe v2's words are a first draft you can correct line by line, not a guarantee.
Can I edit the subtitles after they're generated?
Yes. Click any caption to change its From and To times or its text. Select all captions on the layer to restyle them at once: 10 fonts, size, color, Shadow, Outline or Box, and Top, Middle or Bottom placement. Subtitles don't re-sync on their own if you trim the audio later, so generate them once the cut is locked.
Can I download an SRT or VTT file?
No. Subtitles export burned into the picture, as an MP4 or WebM video at up to 1080p, depending on what your browser can record. There's no subtitle file download and no soft caption track. You can go the other way: upload your own .srt or .vtt, up to 5 MB, and burn it into the video. Uploaded files are added as they are and can't be translated.
Does it show speaker names or sound tags?
No. Scribe v2 labels speakers, but the labels only start a new line when the speaker changes; names never appear on screen. Sound descriptions like [music] or [laughter] are switched off, so the subtitles carry speech only. If you need a name or a tag, click the caption and type it in.
Is it free to start?
Yes. Vidwip is free to start, with no credit card required, and a free start is enough to try it on a short clip. Open the editor, put a clip with speech on the timeline and generate your first subtitles. Importing your own .srt or .vtt, translating, restyling and exporting work the same for every account, and exports carry no watermark.

Make your next video readable on mute

Free to start. Pick a track, get lines timed to every word, and burn them into your video.

Add subtitles to a video