Sign inSign inSign upSign up for freeTalk to sales

Audio translation

Translate audio files into 69 languages, in bulk and across multiple languages at once. Supports MP3, WAV, M4A, OGG, WEBM, and Opus.

Try it for freeSee pricing

Languages

Translate audio into
69 languages

Accurate, context-aware audio translations for global reach. Lara analyses the source track and generates a new translated version, ready for voiceovers, narrations, training material, product videos and podcasts.

View supported languages

Audio formats

6 audio formats,
no conversion

MP3, WAV, M4A, OGG, WEBM, and Opus: upload directly, no conversion needed. A single file can be up to 200 MB in size and up to 2 hours in duration.

Bulk upload

Bulk upload and
multi-language audio

Upload multiple audio files and translate them simultaneously in one session, into several target languages at once. 2 files and 3 target languages produce 6 translated audio files.

  • Add more audio files to the same job
  • Select multiple target languages at once
  • A separate translated audio file for each file and language

Over 500,000 customers trust Lara and Translated for professional translations

  • United Colors of Benetton

Your terminology and style, in every recording

Glossaries, styles and language detection, applied to every file.

Your terminology, enforced on every translation.

Faithful, fluid or creative: you choose, per job.

Lara identifies the source language from the audio file without manual input.

Glossaries

Your terminology, enforced on every translation.

Private by default, ready for your pipelines

Keep recordings out of storage with Incognito mode, automate audio localisation through the API, with enterprise-grade privacy on every plan.

Incognito mode

Files processed in memory and immediately deleted. Zero data retention.

  • Processed in memory
  • Deleted right after delivery
  • Switch it on per session
Incognito mode

API access

Integrate audio localisation into existing pipelines with the REST API and SDKs in 8 languages.

Enterprise-grade privacy

ISO 27001:2022 certified. Data not used to train public models.

  • EU South
  • EU North
  • EU West
  • EU Central

Automate audio localisation via API

Integrate audio translation into existing localisation pipelines with the Lara audio translation API. The process is asynchronous: upload a file, poll for translation status, then download the translated audio.

Read the API docs
  • Rest API
  • SDKs in 8 languages
  • Glossaries & memories
  • Zero retention
{  "filename": "product-demo.mp3",  "target": "it-IT",  "voiceGender": "female",  "style": "fluid",  "adaptTo": ["mem_1234..."],  "glossaries": ["gls_1234..."],  "noTrace": true}

Frequently Asked Questions

Yes. The Lara Translate Audio tab supports both bulk file translation and multi-language translation in the same session. Upload multiple files using the Add more audio files button, then select multiple target languages. Lara processes all combinations simultaneously and returns a separate translated audio file for each file and language pair. For example, 2 files and 3 target languages produces 6 translated audio files. Files are available for download for 1 hour after the job completes.

Lara Translate accepts six audio formats: MP3, WAV, M4A, OGG, WEBM, and Opus. There is no need to convert your file before uploading.

Lara Translate supports audio translation in 69 languages, including English, Spanish, French, German, Italian, Portuguese, Japanese, Korean, Chinese, Arabic, and many others. The full list of supported language codes is available in the API documentation.

Usage is calculated at 40 characters per second of audio. A 15-second voice note uses 600 characters (15 × 40). Usage is drawn from your shared character pool, which covers text, document, image, and audio translation. For a full breakdown, see How is usage calculated?

Yes. A single audio file can be up to 200 MB in size and up to 2 hours in duration. If your file exceeds either limit, split it into smaller segments before uploading.

Translated files are available for download for 1 hour after the job completes. After that window closes, files are removed from the system. Enable Incognito mode if you want files processed entirely in memory with zero retention from the start.

Yes. The Lara audio translation API supports automated localisation workflows. The process is asynchronous: upload a file, poll for translation status, then download the translated audio. You can apply glossaries, translation memories, and style preferences. SDKs are available for Python, TypeScript, Java, PHP, Go, Kotlin, C#, and Swift.

Ready to translate like the world’s best teams?

Start free in seconds. Upgrade only when, and as much as, you actually translate.

Sign up for freeTalk to sales