How to translate audio to text
It's two jobs wearing one name — transcription, then translation. Get the order right and the result is auditable.
By Anubhav Jain, Trustample founder · Updated July 2026 · Claims verified against the live product
Free: 60 min per 30 days, files up to 30 min / 50 MB · Paid plans: 10 to 100 hrs, files up to 2 GB. Big video? Upload just its audio track — same transcript, much faster upload.
Encrypted in transit · audio & video auto-delete in 30 days · never used to train AI
Optional settings
Set this for the most accurate result on quiet or mixed audio.
Get a second copy translated into another language (Pro).
Uncommon words the AI might misspell — list them so they come out right.
"Translate audio to text" bundles two different jobs. First, transcription turns the spoken words into text in their original language. Then translation turns that text into the language you want. Every reliable workflow does them in that order — because you can only translate text, and audio isn't text until something transcribes it.
The fastest path runs both in one upload: choose the spoken language, switch on translation, and get the original transcript plus its translation side by side. Trustample does this on Pro — a Japanese, Russian, Spanish or French recording comes back as both the native transcript and an English version you can check against it. Keeping the transcript visible is the whole point: a misheard word shows up as a fixable error in the original, instead of hiding inside the translation where no one can see it.
The free path splits the two steps: transcribe on Trustample's free plan (60 minutes a month, no card), export the text, and paste it into any free text translator. You give up the stored side-by-side view and subtitle timing, but for personal use it's a real answer. And a word on "can Google Translate do audio to text?" — its microphone and conversation modes translate live speech in short bursts well, but they aren't built to take an audio or video file and hand you a saved, editable, translated transcript. For a file, transcribe-then-translate is the route.
How it works
- 1Upload the recording and set the spoken language (naming it beats auto-detect on tricky audio).
- 2On Pro, set 'Translate transcript to' — or on free, just transcribe and export the text.
- 3Verify the original against the audio, then translate; the translation inherits your corrections.
Frequently Asked Questions (FAQs)
Can I translate audio to text for free?
Yes, in two moves: transcribe free (60 minutes a month, TXT export, no card), then run the exported text through any free text translator. You skip the stored translation and subtitle exports, but nothing about the core job costs money for light use.
Can Google Translate turn an audio file into text?
Not a file directly — its strengths are live microphone input and short back-and-forth conversation mode, not ingesting a recording and returning a saved transcript. To translate a file's speech, transcribe it first, then translate the text; that also gives you something you can edit, search and turn into subtitles.
Why not use a tool that translates speech in one step?
Some do — but the transcript still exists inside the pipeline; hiding it just means you can't check it. A misheard name or figure becomes an invisible translation error. Keeping the original transcript visible makes the output auditable, which matters the moment the text leaves your screen.
How accurate is audio-to-text translation?
Two error rates stack: transcription (roughly 95%+ on clear audio in major languages, less on noise or heavy dialect) and machine translation (strong on factual speech, weaker on idioms, sarcasm and jargon). Correct the transcript before translating, and have a fluent speaker skim anything public-facing.
Which languages can I translate to and from?
Transcription supports 16 spoken languages including Japanese, Russian, Spanish, French, German, Portuguese, Chinese, Korean and Arabic; on Pro the transcript translates into English (and several other targets). The language-specific guides — like Japanese video to English — walk through each one.
An honest note: Machine translation, one-step or two-step, produces a working draft — never treat it as a certified translation for legal, medical or broadcast use without a human review. The two-step workflow's advantage isn't perfection; it's that the draft is auditable, because you can see and fix the words it started from.
Related pages
The workflow, step by step
Russian Video to EnglishSame two steps, another language
Can Google Translate a Video?The video version of this question
Explore everything in Answers or browse all transcription tools.