Translate YouTube videos without subtitles

A useful video does not always have subtitles in your language. VGLOT prepares audio in 60-second playback intervals, then puts the original words beside a translation in a panel synchronized to playback.

Last updated on By Knoveya4 min read
Follow:
Recorded VGLOT popup with English source, Vietnamese target and Start captions button
Choose Source and Target before Start. This recorded screen may differ from the current Store release; version 0.4.0 prepares playback intervals before their captions appear.

Get started

  1. Add VGLOT to Chrome from the Chrome Web Store. Open the extension popup and Connect with Google using your Knoveya account.
  2. Start Free with 30 caption minutes per UTC calendar month, shared between extension and web usage. Standard and Premium offer larger monthly allowances; no paid plan is required to start.
  3. Open a public YouTube watch page. Choose from 61 explicit Source options plus Auto detect and 109 Target options. Explicit Source guides recognition; Auto is limited to supported detection languages. These expanded choices apply to the extension only.
  4. Press Start captions at the playback position you want. VGLOT reuses ready captions for your account and exact language pair, follows accepted processing, or prepares the current 60-second interval. Wait for that interval to be ready.
  5. Read both languages in the synchronized panel. While playing, VGLOT may prepare one next interval. Pause stops new prefetch. Stop prevents new requests; accepted work may finish. Hiding the panel does not stop processing. Drag, resize or adjust text appearance, and click a timestamp to revisit a moment.

What happens when a subtitle track is missing?

VGLOT uses the public video’s audio to create timestamped source text and a translation. It does not require an existing subtitle track. The two languages appear in a movable transcript panel rather than a dubbed audio track.

This helps with tutorials, interviews and lectures whose speech you can hear but cannot comfortably understand. It does not recover silent speech, text shown only in slides or words hidden by loud music. A usable native subtitle track remains another option.

Wait for processing, then check important wording

VGLOT prepares the current 60-second interval before displaying its captions. While playing it can prefetch one following interval; pause stops new prefetch. Waiting depends on the audio, language and service availability. There is no live transcription or sentence-by-sentence fallback. Ready intervals are reused.

Names, accents, overlapping speakers and specialist vocabulary can cause mistakes. Compare the original and translated lines, replay the audio and verify important facts against a reliable source. Translation is a reading aid, not an authoritative transcript.

Account, Free access and quota

Connect with Google. Free includes 30 minutes per UTC calendar month, shared with web processing. Standard includes 400 and Premium 1,000 minutes by default; legacy Plus maps to Standard.

Newly completed extension intervals count toward quota, including completed prefetch and silent intervals. Each 60-second interval costs one minute; a shorter final interval rounds up to one minute. Failed processing is refunded. Ready captions remain reusable after exhaustion, without another processing debit. The web workspace still uses whole-video processing quota.

Caption caches belong to your account and exact language pair. A ready English-to-Vietnamese pair does not automatically provide English-to-Japanese captions or make them available to another account. Auto and explicit Source choices are separate cache keys.

Where the audio and text go

After Start, Knoveya temporarily processes public video audio with Deepgram for transcription and sends transcript text to Gemini for translation. Extension intervals are cached privately for your account, video and exact Source/Target pair. Temporary audio is deleted after processing. Extension-only intervals do not create partial Library entries; the web workspace still processes complete videos into your private Library.

The extension does not capture tab audio, use your microphone, or perform live speech recognition. Language and appearance preferences stay on your device. Review the privacy details before processing material you consider sensitive.

If captions do not appear

Check your connected account, selected languages, remaining quota and connection. The popup shows processing status or an error. If a job fails, resolve the cause and use Retry captions; failure does not start live capture.

Only public YouTube watch pages in Chrome are supported. Shorts, livestreams and private videos are not supported; restricted or unavailable videos can also fail. Contact VGLOT support with a description of the issue without including private transcripts or credentials.

Related guides