Translate Audio to English Text
Two checked steps: transcribe the original language and review it, then create written English beside it in the same saved task. Not dubbing or voice-over.
Whisper Web
0 min left
Sign in to create a saved task. If your file is unavailable after signing in, select it again.
Dashboard
How do you want to transcribe?
Free minutes are included. Upload a file or record audio to start.
Sign in to submit. Transcription uses account credits; Translate to English needs a Pro or Max plan and uses additional credits. See plans and usage limits. Audio is uploaded for cloud processing.
How to Translate Audio to English Text
- 1
Transcribe the Original Language
Upload a file or record speech. Leave language detection on, or select the spoken language when you know it. Check the duration and credit estimate, then sign in to submit. The first result is a transcript in the language that was spoken, not English.
- 2
Review the Source Transcript
Listen back and correct names, numbers, and specialist terms in the original language. A mistake fixed here does not travel into the English text. Saved edits create a new transcript revision on the same task.
- 3
Translate to English
Choose Translate to English. The saved transcript is sent for written translation and the English appears beside the source. Copy it or download a TXT file. If the step fails, retry translation only; transcription does not run again.
Examples: Chinese and Spanish Audio to English
Mandarin product meeting → English follow-up
A team lead records a Mandarin sync and needs English notes for a regional manager. Check the romanized name and the delivery date in the Chinese transcript first, then create English text for the same task.
- Chinese source
- 客户反馈的导出问题，由王磊负责跟进。
- English text
- Wang Lei will follow up on the export issue customers reported.
Spanish customer interview → English research notes
A researcher interviews a customer in Spanish and shares findings with an English-speaking product team. Confirm industry terms such as proveedores in the source before relying on the English wording.
- Spanish source
- Lo que más nos costó fue exportar los informes al formato de contabilidad.
- English text
- What we struggled with most was exporting reports to the accounting format.
Authored illustrations with original example sentences, not customer recordings or measured accuracy results. Your recording needs its own review.
Why the Original Transcript Comes First
When you translate audio to English text in one opaque step, a misheard word becomes a confident English sentence and you have nothing to check it against. This page keeps the original-language transcript and the English result in the same saved task, so a bilingual reviewer can compare them line by line.
The English text records which transcript revision it was created from. If you edit the source after translating, the English is marked out of date instead of silently mixing old and new wording. Translate the updated transcript when you are ready; the previous English stays visible, greyed out, until then.
If the English step fails, the source transcript remains saved, credits for that failed attempt are returned, and Retry English translation repeats only that step. When the detected language is already English, the page says so and does not offer a translation you do not need.
Audio to English Text vs Other Language Workflows
| Need | Use | What you get |
|---|---|---|
| English is the only target | This page | Source transcript beside English text, revision check, English TXT download. |
| Several target languages | Audio Translation Studio | Choose Spanish, French, Japanese, or another written target for a reviewed transcript. |
| Original-language text only | Speech to Text AI workspace | Transcript, subtitles, summaries, and exports without a translation step. |
Some speech models can output English directly from foreign-language audio; the OpenAI Whisper repository documents a translate task for this. That shortcut skips the reviewable source text. This page deliberately produces the original transcript first, then translates the reviewed text.
Supported Files, Languages and Plan Limits
Upload MP3, WAV, M4A, AAC, FLAC, OGG, OPUS, WebM, MP4, or MOV, or record from your microphone. The shared cloud service accepts files up to 1 GB and 60 minutes. Mixed-language speech, heavy accents, and background noise need closer review in both columns.
Transcription uses account credits. Translate to English is an AI action that requires an active Pro or Max plan and uses additional credits, shown against your balance. Compare plans and usage limits before starting. Opening a saved task again does not charge for the transcript or an English result that is still current.
The output is written English text. It is not dubbed audio, a cloned voice, or burned-in subtitles. Treat machine translation as a working draft; contracts, medical or legal material, and publication copy need qualified human review.
Frequently Asked Questions
Which languages can I translate into English?
Speech recognition supports many languages, including Chinese, Japanese, Spanish, Korean, French, German, and Portuguese. Quality varies with the language, audio, and vocabulary. Try a short representative recording and review the source transcript before translating longer material.
What happens if my recording is already in English?
When the detected or selected language is English, the transcript is already your English text. The page explains that no translation is needed and does not charge translation credits. Edit and export the transcript from the Original or Edited tab.
I edited the source transcript. Is the English updated automatically?
No. The English keeps the wording it was created from and is marked out of date, so you can see what changed. Choose Translate updated transcript to create English for the current revision. Nothing is re-translated without your action.
The English translation failed. Do I need to upload again?
No. The original-language transcript stays saved. Credits for the failed translation attempt are returned. Select Retry English translation on the same task; the recording is not transcribed again.
Can I keep both the source and the English?
Yes. Both belong to one task in Recordings, and the task link reopens this page. Download the English as TXT, and export the original-language transcript as TXT or DOCX from the transcript tabs.