InsTranscript
…Credits

AI Subtitle Translator

Translate timed subtitles generated from a supported video, preserve cue timing, and export SRT or WebVTT beside the original transcript.

Public social video link

Paste a public Instagram video link. All six platforms work here.

Example preview

Transcript

English 00:5887 words

00:00Hey everyone! Today I'm making a super easy garlic lemon pasta.
00:08First, bring a pot of salted water to a boil.
00:14Add your pasta and cook until al dente.
00:20While that cooks, heat olive oil in a large pan.
00:26Add minced garlic and cook for about thirty seconds, until it smells amazing.
00:33Then add cherry tomatoes and let them soften.
00:42Toss in the cooked pasta, a squeeze of lemon juice, parmesan, and fresh basil.
00:52Give it a good mix, and that's it! So simple and so good.
Illustrative sample · not a live transcription
Preview first, then sign in and claim 500 trial credits in eligible regions. One link at a time; paid plans support up to 50.
Upload video or audio

MP4, MOV, WebM, MP3, WAV, M4A, FLAC, OGG or AAC · up to 25 MB. Your balance and the configured duration limit apply. Files are private; 1 credit per second, rounded up.

1 credit per second, rounded up. Minimum 30 credits per video. Failed transcription returns reserved credits.

Sign in to continue

  1. Paste link

    One public video link

  2. Preview video

    Check the source clip

  3. Transcribe

    Convert speech to text

  4. Get text

    Copy, TXT or timed SRT

Generate subtitles, then translate them

Use a supported video link or upload audio or video to obtain a transcript with genuine timing. This subtitle translator translates subtitles generated by this tool; direct SRT or WebVTT file import is not currently supported. You can choose an output language before transcription or select Translate after the original result appears. Existing YouTube captions can provide a timed source without speech recognition credits.

Preserve each subtitle cue

A translated cue inherits its original start and end time. The model translates text, while the server verifies that every source segment has exactly one output entry. Missing, repeated or unknown segment indexes are rejected. This preserves the relationship between subtitles and source playback without pretending that translation can improve inaccurate speech recognition timing. The original subtitles always remain available in a separate tab.

Choose SRT or WebVTT

SRT is selected by default for genuinely timed results. Select WebVTT for a browser-compatible subtitle file, or TXT when you need readable text without cue numbers. Export uses the currently active original or translated version. If the source has no reliable timing, the tool offers TXT and disables subtitle formats instead of manufacturing timestamps. Preview the output in your editing software and adjust long translated lines if needed.

Translation access and quality

Sign in with a verified account to use the 20-language selector. Included translation is limited to 16,000 characters and 200 source segments, with at most two provider calls per transcript and daily usage caps. Completed translations are cached for repeat viewing. Language selection is not a guarantee of perfect localization: idioms, proper names, technical vocabulary and timing-sensitive captions still need human review.

A trial without automatic spending

Opening the page does not create credits or start speech recognition. Preview your source, sign in, then claim a one-time 500-credit trial in an eligible region. Before audio transcription, the server measures duration and reserves credits at 1 credit per second, with a 30-credit minimum. Translation failures leave the original intact. Paid subscriptions and packs provide permanent accumulating credits, and paid users can process up to 50 links in a batch.