Start free

TikTok transcription.Carava link, raici captions.

Carava TikTok video URL. Ni vakavakavidi kina ira na audio server-side me vakabubutaka timestamped text me SRT me VTT caption files — lawa na me vakarau-vuraqa o ni cauravou.

Drop a file, or pick one

MP3 · WAV · M4A · MP4 · MOV · MKV · OGG · OPUS · FLAC · WEBM — up to 100 MB anonymously

Paste a link, we’ll fetch the audio

YouTube · TikTok · Vimeo · Twitter · SoundCloud · Spotify · 50+ more

Record straight from your browser

Sign up takes 30 seconds — recording opens right after, in the dashboard.

No card required~90s per 60-min fileSRT · VTT · DOCX · TXTFiles auto-deleted in 24h

↓ Vosa vinaka ira na output

Public URL mai. Captions mai.

Carava public TikTok video link. Ni vakavakavidi kina ira na audio track, run language detection, stream back captions while background music keep playing under voice.

TikTok video URLREC 1 voice · 0:47 · vertical 9:16
auto-detected en-US44.1 kHz · music bed -18 dB
~90s
Captions · streaming94% accuracy
S1

Okay kena na secret na crispy tofu walang iko vakatau — vakasama kena ten minutes, walang two.

S1

Drega cornstarch, walang flour. Toss kena, walang dust kena.

S1

Air fryer at 400 para twelve minutes, flip halfway.

S1

Comment 'tofu' me vakatuwatuwai kina na full sauce recipe.

94% on creator voice-overSRT · VTT · TXT · DOCX · JSON

↓ This is the dashboard

This is what loads when the job finishes.

Same layout as the real dashboard — Summary, full Transcript, Speakers tab, Exports. Key points and action items extracted automatically. Auto-tags on every job.

Try it on your own file — it's free

Three real options · honest comparison

TikTok auto-captions. CapCut or Submagic. O keda.

TikTok ni ships auto-captions sa editor. CapCut me Submagic ni add styled, animated captions para re-upload. Ni vakavidi kina raw transcript me clean SRT/VTT — bring your own editor.

Option 01

TikTok auto-captions

Built into TikTok editor. Toggle on, captions appear. Walang file o moce rawa mai elsewhere.

RequiresUpload through TikTok app
Language coverage~40 languages, EN strongest
ExportNone — burned in only
Edit before publishIn-app text editor
Music handlingMisses lyrics, garbles voice over loud beds
CostFree
Best forCreators wedru captions inside TikTok walang vakarau-vuraqa sa Reels o Shorts.
Option 02

Transcription.Solutions

Carava public URL. Tikivu na transcript file me SRT/VTT o rawa moa ni vakasama sa any editor o vakarau-vuraqa anywhere.

RequiresPublic TikTok URL — no login
Language coverage100+ with auto-detect
ExportSRT · VTT · DOCX · TXT · JSON
Edit before publishWeb editor, drega re-export
Music handlingVoice isolation on noisy beds
Cost · per min$0.03
Best forCreators cross-posting sa Reels/Shorts/YouTube, agencies repurposing client TikToks, researchers archiving trends.
Option 03

CapCut / Submagic

Styled, animated captions tuned para short-form. Locked sa editor, English-first.

RequiresApp install + paid para export
Language coverage~20 strong, others spotty
ExportMP4 with burn-in, SRT on paid
Edit before publishInside timeline only
Music handlingEN-tuned, drops on accented voice
Cost$10–24/mo (approximate, 2026)
Best forSolo creators wedru animated word-pop captions me walang vei leave CapCut/Submagic editor.

Pricing approximate as of May 2026. Language counts based on each vendor's published support pages.

Specific to TikTok

Three things na bite people on generic transcription tools.

TikTok audio walang podcast audio. Ira na differences ni worth flip before queue job.

What goes wrong

  1. 1Background music get transcribed as speech. Generic ASR hear lyrics me write alongside voice — caption file become unusable.
  2. 2Creator slang me handles (@username, 'rizz', 'fanum tax', product names) come back phonetically misspelled o auto-corrected sa wrong word.
  3. 3Fast hooks — first three seconds where creators stack 15 words beat swipe — get clipped o compressed because ASR still warming up.

What to flip here

  1. 1Turn on Voice isolation on job form. Separate voice stem from music before transcribing, so trending audio walang pollute captions.
  2. 2Paste handles, brand names, me creator-specific vocab into Custom vocabulary. Pass as recognizer hint — case me spelling come back correct.
  3. 3Set Caption format sa short-form (max 3 words per line, 1.2 sec per cue). SRT come out pre-formatted para vertical video walang manual line breaks.

Recommended job settings para TikTok

Carava TikTok URL me ira na flip on by default. Override per-job from form.

Source
Public URL · audio extracted server-side
Voice isolation
On (music bed suppressed)
Language
Auto-detect · 100+ supported
Caption format
Short-form · 3 words/line · 1.2s cues
Filler words
Kept (creators rely on ira)
Export
SRT · VTT · TXT · DOCX

Accuracy · real-world numbers

94% on clean voice-over. Music-heavy clips drop predictably.

Ceiling ni set by how loud music bed me how fast creator talk. Voice-over recorded separately me dropped over quiet bed ni best case; lip-sync trends me duets ni worst. Numbers below mai from real TikTok URLs run through pipeline.

94%
Voice-over · quiet music bed

Creator recorded on mic, music sit 15-20 dB below voice. Talking-head educational me recipe content land here.

91%
On-camera · phone mic · no music

Selfie-style talking head, no backing track. Phone mic me room reverb cost few points versus voice-over.

85%
Loud trending audio under voice

Voice me music within 6 dB. Fast hooks me brand names take hits — expect 1-minute clean-up pass.

78%
Duets, stitches, lip-sync clips

Two audio tracks overlapping o song lyrics mouthed. Transcribe na speak; song lyrics flagged, not retyped.

Common questions

8 things people ask about TikTok transcription.

01Do I need to download TikTok first?+
No. Carava public video URL (share link from TikTok app) me extract audio server-side. If video private o region-blocked, download MP4 yourself me upload — walang bypass TikTok access rules.
02Will you transcribe song lyrics o just creator voice?+
Just spoken voice. Voice isolation suppress music bed before transcription, me trending-audio lyrics get flagged in JSON output rather than write sa caption track. Flip isolation off if specifically want lyrics.
03Can I get SRT formatted para vertical short-form video?+
Yes. Short-form caption preset break cues at roughly 3 words per line me 1.2 seconds per cue — rhythm na fit 9:16 safe zone walang overlap UI. Standard SRT (one sentence per cue) also available.
04What about duets me stitches with two voices?+
Acoustic diarization separate two voices me label Speaker 1 me Speaker 2. Accuracy drop 5-10 points when audio tracks overlap heavy — worst case in data.
05Does it handle non-English creators?+
Yes — 100+ languages with auto-detect. Spanish, Portuguese, Indonesian, Vietnamese, me Arabic creators come back at roughly same accuracy band as English. Code-switching (mixing two languages mid-sentence) detect me label per segment.
06How long until transcript ready?+
Under five minutes para standard 30-90 second TikTok, usually under two. Longer-form TikToks (3-10 minutes) finish ni roughly 1/10 of real-time.
07Can I bulk-process creator's whole feed?+
Yes, via API o by pasting list of URLs into dashboard. Rate-limit URL fetcher politely so TikTok walang block — expect ~30 videos in first batch, drega steady throughput.
08Is this allowed under TikTok's terms?+
Fetch public videos via public share endpoints — same way browser preview does. Walang bypass private accounts o login walls. If transcribing someone else's content para commercial use, fair-use me platform rules on you check.

Carava TikTok URL. Vosa output.

30 free minutes every month. No card. SRT, VTT, 100+ languages, all exports included.

Start free