Paste an Arabic video or audio link and get a timestamped transcript in Arabic script. Pulling existing captions is free and needs no account. Where there are none, AI transcription handles the audio, in Modern Standard Arabic or regional dialect.
Open JotVox →Paste a link from TikTok, YouTube, Instagram, Facebook, X or another supported platform, or upload a file from your device, Google Drive or Dropbox. JotVox checks for captions and pulls them when they exist: free, no sign-up, timestamps included.
When there are no captions, AI transcription takes over, transcribing the audio at 98%+ accuracy on clear speech for one credit. Three free credits come with signing up, and a verified email adds one free AI transcription a day. Either route gives a timestamped Arabic transcript, exportable as TXT, SRT, VTT or MP3, with optional AI summaries.
Arabic transcription has one dominant difficulty: the Arabic people write is not the Arabic people speak. Modern Standard Arabic is the variety of news bulletins, formal speeches and written material. Conversation happens in regional dialects, and distant regions can find each other hard to follow.
Egyptian Arabic travels furthest, carried by decades of Egyptian film and television. Levantine covers Syria, Lebanon, Jordan and Palestine, and Gulf Arabic spans the peninsula. Maghrebi Arabic, across Morocco, Algeria and Tunisia, is the most divergent of the common groups and generally the hardest to transcribe, on vocabulary and because speakers mix in a lot of French. Gulf and Levantine speakers code-switch into English the same way.
A subtler problem sits underneath. Modern Standard Arabic has a settled spelling; the spoken dialects do not, so a transcript of dialect speech has to pick a convention for words rarely written down. Expect a news read to transcribe more cleanly than a fast dialect vlog.
SRT and VTT files are plain UTF-8 text. Arabic inside them is stored in logical order, the order the words are spoken, and the software displaying the file decides the visual order using the Unicode bidirectional algorithm. That split between stored order and displayed order is behind nearly every right-to-left subtitle complaint.
What you notice first is punctuation and numbers appearing at the wrong end of a line. Cue numbers, timestamps and the --> arrow are left-to-right characters sitting beside right-to-left text, so an editor with weak bidirectional support can place them confusingly, and a sentence-final full stop can show up on the left. Lines mixing Arabic with an English brand name or a figure look worst.
In almost every case the file is correct and only the display is wrong. Check it in the player you will actually use, not a basic text editor. Do not fix the order by retyping the characters backwards: that corrupts the text, and it will then be wrong everywhere, including in players that were rendering it correctly.
Two more things. Keep the file as UTF-8: opening one in a tool that assumes an older Arabic code page produces mojibake. And Arabic letters change shape according to their neighbours, so software without Arabic shaping shows them detached. Both are reader-side problems, fixed by using software with real Arabic support rather than by editing the transcript. SRT to TXT strips the timing without touching the Arabic.
Arabic speech comes back in Arabic script, without diacritics. Short vowel marks are normally left off written Arabic outside teaching material and the Quran, so an undiacritised transcript is ordinary output, not missing information.
JotVox transcribes, it does not translate. An Arabic video produces an Arabic transcript. For English, export the TXT and translate it with a separate tool. For other languages, Audio to Text covers all 99 that AI transcription supports.
TikTok and Instagram Reels carry most of the Arabic short-form video people want in text, much of it dialect speech recorded on a phone over a music bed. YouTube holds the longer material: news segments and interviews in Modern Standard Arabic, religious lectures and long-form podcasts. Facebook matters for news pages and Live replays, X for fast news clips.
Music over speech is the usual culprit, and Arabic short-form video leans on backing tracks heavily. Also hard: panel debates where several people speak at once, which can give accurate words with the wrong speaker attached; phone audio with wind, street noise or room echo; dialect-heavy speech, with Maghrebi the most demanding; and sung or chanted delivery, which is harder than plain speaking. Private, deleted, age-restricted and region-locked links cannot be fetched at all; where you have the file, upload it directly.
1. Paste the link or upload the file Copy the Arabic video URL from TikTok, YouTube, Instagram, Facebook, X or another supported platform and paste it into JotVox, or upload an audio or video file from your device, Google Drive or Dropbox.
2. Pull captions or run AI transcription Existing captions are pulled free with no sign-up. When a video has none, AI transcription writes the transcript from the audio for one credit, handling Modern Standard Arabic and regional dialect speech.
3. Read the Arabic transcript The transcript arrives in Arabic script with timestamps, undiacritised as written Arabic normally is. Use the timestamps to jump to the part of the video you need.
4. Export, and check right-to-left rendering Export as TXT, SRT, VTT or MP3. Open subtitle files in the player or editor you will actually use, since basic text editors often display right-to-left text in a confusing order even when the file is correct.
No. JotVox transcribes and does not translate, so an Arabic video gives you an Arabic transcript in Arabic script. To read it in English, export the TXT and run that text through a translation tool. Doing it in that order is usually better anyway, because you can correct names and dialect spellings in the transcript before translating.
Both. AI transcription works on dialect speech as well as Modern Standard Arabic, which matters because almost all conversational Arabic video is dialect. Accuracy is generally highest on formal Modern Standard delivery, such as news reads, and lowest on fast, heavily regional speech. Maghrebi Arabic is usually the most demanding, partly because of extensive French code-switching.
Almost always the file is correct and the display is not. Arabic is stored in logical order and reordered for display by the bidirectional algorithm, so timestamps, numbers and punctuation can appear at an unexpected end of the line in editors with weak right-to-left support. Open it in your actual player, keep it as UTF-8, and never retype the characters backwards.
Pulling existing captions is free with no sign-up. AI transcription, needed whenever a video has no captions, uses credits: three free credits on signup, then one free AI transcription a day once your email is verified. After that, credit packs start at $4.99 for 20 credits and subscriptions start at $9 a month. One credit covers one action.
No, and that is normal. Arabic is written without short vowel marks outside teaching texts, dictionaries and the Quran, so transcripts come back undiacritised like ordinary Arabic writing. Nothing is lost: the consonantal text is what Arabic readers expect. If you need diacritics for a specific purpose, add them in a dedicated Arabic text tool after exporting.