Smartcat transcribes the speech in your MP3, lets you correct the transcript once, translates it into any of 280+ languages, and exports the result as text, subtitles, or — in 35 voice-over locales — an AI voice track. Free for 15 days with 15,000 Smartwords, no credit card.
1,000+ enterprise brands trust Smartcat to translate recorded content
To translate an MP3 file, upload it to Smartcat exactly as it is — no conversion to text first. Smartcat transcribes the speech into a timed, speaker-labeled transcript; you review it once and fix any misheard names before anything is translated. The AI then translates the corrected transcript into any of 280+ languages, reusing the glossary and translation memory attached to your project so terminology stays consistent.
You export the result as a translated document, a subtitle file (SRT or VTT), or an AI voice track in one of 35 voice-over locales. Pricing is metered at one Smartword per source word of transcript — not by recording length — so a long recording with little speech costs less than a short, densely scripted one.
What matters is the recording, not the encoding: a file that's muddy to your ear will be muddy to the model too. Bitrate and encoding make no difference — any bitrate, constant or variable, uploads the same.
1
Upload the MP3
Drop in one file or a whole folder — there is no cap on files per project, and batch uploads of hundreds of files work the same on every plan.
The constraint is per-file size: up to 512 MB — split a longer recording into parts. No published limit on recording length.
2
Check the transcript
Smartcat transcribes with timestamps and speaker labels. A misheard product name or acronym gets fixed here, once — before it can be translated into every target language.
3
Pick languages and output
Translated text or subtitles in any of 280+ languages, or an AI voice track in one of 35 voice-over locales. English out, English in — the same flow translates to English or from it.
4
Export — or send to review first
Download the result, or route any language to a vetted human reviewer inside the same workflow. Batch job? Select the projects and download once — one ZIP, every file in its original format, sorted into folders.
No conversion first, no transcript to prepare — your MP3 uploads exactly as it is. Correct a name once in the transcript, and every target language comes out right.
280+
Languages for text and subtitles
Your MP3 can come back as a translated document or an SRT/VTT subtitle file in any of Smartcat's 280+ languages, in either direction.
35
AI voice-over locales
Synthetic speech output is a narrower set than text: 35 locales, powered by ElevenLabs. Where a locale has no voice, export text or subtitles instead.
1
Smartword per source word
An MP3 is metered by transcript word count, not minutes — a 60-minute walkthrough with sparse narration can cost less than a 10-minute scripted ad. AI dubbing is metered at ten Smartwords per word. Details on the pricing page.
One upload can come back four ways: a translated transcript (a timestamped document per language), subtitles (SRT or VTT timed to the original speech), an AI voice track in one of 35 locales, or any of those checked by a human reviewer before export.
Three things break recordings before translation even starts — and none of them is the translation itself. Knowing them up front saves a re-upload.
Noisy or muddy audio
Speech-to-text quality tracks recording quality. Background music, wind, and heavy compression degrade transcription — and an error made there is translated faithfully into every language. Check the transcript before you trust the output.
Songs and lyrics
Sung vocals with instruments behind them transcribe far worse than speech. This is a speech tool — for lyric sheets you already have as text, translate the text file instead.
Everyone talking at once
Crosstalk on a panel or a heated call gets merged or misattributed. Speaker labels handle turn-taking well; simultaneous speech needs a pass in the editor.
The transcript editor shows you exactly what the model heard — try it on a noisy file before you commit to anything.
2–3 days
Turnaround, Down From 10
Smith+Nephew cut eLearning translation turnaround from an average of 10 days with their previous providers to two to three days for the same course length — recorded and video material included.
70%
Less Reviewer Workload
Smith+Nephew's reviewers spend 70% less time editing — corrections made once carry into every language via translation memory.
+30%
More Content Output
Wunderman Thompson increased content output by 30% with Smartcat across its production workflows.
Listen to your source first — clear, single-speaker audio transcribes best. Split recordings larger than 512 MB into parts. Load your glossary before you start, so recurring names come out identical — then fix the transcript once, and the fix carries into every language.
AI Audio Translator
Any other audio format — M4A, AAC, OGG, FLAC, WMA — or you're not sure what the file is. The head page for everything recorded.
AI Video Translator
Video with an audio track — subtitles, dubbing, and voice-over. Speech only; text burned into the picture has to be changed in the source.
SRT Translation
Already have a subtitle file? Translate the .srt or .vtt directly, keeping every cue and timestamp.
AI Dubbing
Voice replacement on video, in any of the 35 AI voice-over locales.
Upload the file, fix the transcript once, and export text, subtitles, or speech in as many languages as you need. Free for 15 days with 15,000 Smartwords — no credit card.
Smartcat is free for 15 days with 15,000 Smartwords and no credit card — full access to translation capabilities, enough for several hours of typical speech. After that, you pay per Smartword, not per file or per minute.
No — it runs in the browser. “Download” happens at the other end: you download the translated output (document, SRT/VTT, or audio file). Nothing to install.
No — transcription is the first step of the product. Drop the MP3 in and Smartcat produces a timed, speaker-labeled transcript automatically; you review it once, then translate.
MP3 is natively supported, alongside MP2/M2A, M4A, AAC, OGG, FLAC, and WMA — files up to 512 MB each; split a longer recording into parts. Bitrate and encoding (CBR or VBR) make no difference. There's no published limit on recording length and no cap on files per project; the ceiling is per-file size, the same on every plan. If your audio is inside a video container, use the AI video translator instead.
Yes — from any of 280+ source languages, and the same flow works in reverse: take an English recording into every market language at once.
You choose: text, subtitles, or an AI voice track in one of 35 voice-over locales. Voice output is a narrower set than the 280+ languages available for text.
Accuracy is set at the transcription stage, and transcription tracks recording quality. Clear single-speaker speech comes out well; heavy background noise or crosstalk needs a pass in the transcript editor. That's why the workflow puts the editable transcript before translation.
Yes — route any language to a vetted human reviewer from Smartcat's Marketplace inside the same workflow, and their corrections train your translation memory.
Question not answered here? Book a demo — a 1:1 consultation, no commitment.
Smartcat Help Center. Translate video, audio and transcripts.
Smartcat. How Smartcat AI enabled Smith+Nephew to cut eLearning translation turnaround. Smith+Nephew case study.
Smartcat. Wunderman Thompson case study.
Smartcat. Pricing.