# Speech to Text

Turn a recording into a transcript you can review, correct and export.

> Canonical: https://aimusicis.com/speech-to-text

## Recording and transcript limits

### 15 minutes

Maximum duration of the recording submitted for transcription.

### 50 MB

Maximum uploaded audio file size. Choose a shorter recording if it exceeds either limit.

### 4 export formats

Download readable text, SRT or VTT subtitles, or structured JSON with available timing information.

## Use speech to text without losing the timing

### Choose the recording

Pick a completed track from your music library, upload a playable audio file, or use the recording tab. Preview the source and check that the voices are audible before submitting.

### Transcribe and return to the result

Leave the language on Detect automatically, or choose one of the 100 languages the model transcribes. Review the credit amount and start transcription. My transcripts keeps earlier tasks available, including tasks that are still processing.

### Correct words and save a copy

Review the transcript in groups of 20 segments. Each segment shows its start and end time, and its play button plays the recording from that point so you can check a word against what was said. Correct recognition errors, then copy the text or export the format your destination editor accepts.

## Tools related to Speech to Text

### [AI Text to Speech](https://aimusicis.com/text-to-speech)

### [AI Music Generator](https://aimusicis.com/ai-music-generator)

### [Free AI Beat Generator](https://aimusicis.com/ai-beat-generator)

### [AI Vocal Remover](https://aimusicis.com/ai-vocal-remover)

## Speech to text workflow questions

### Which recordings work best for speech to text?

Clear speech with limited background sound gives the model a more useful source. Overlapping speakers, loud music and indistinct words can reduce accuracy. Listen to the source and proofread names, numbers and technical terms in the transcript.

### Are timestamps always available for every word?

Speech to text returns timed segments and retains word alignment when the provider supplies usable timing. Some segments may only have a start and end time. Correcting a segment's words removes that segment's old word alignment while retaining its segment timing.

### Can the transcript name speakers and note sounds?

Yes. Choose Identify speakers automatically, or a count from 2 to 6, and each line is labeled with who is speaking. Turn on Include sound notes to mark sounds such as [laughter] or [applause] where they happen; with sound notes on, speakers are found automatically. Review recordings where voices overlap carefully before using the transcript.

### Which export should I choose?

Use TXT for notes or a document, SRT for subtitle editors, and VTT for web captions. JSON keeps the language, text, timed segments and any remaining word alignment for software that can read structured data.

### Are my corrections saved in the original task?

Corrections remain in the current editor and are included in copied text and downloads. They do not overwrite the original transcription in history. Export your revised copy before leaving the page.

### Can I use a transcript in a music video?

The music-video editor uses the same transcription and review controls. Apply a transcript to that video's audio to replace its caption timeline, then review the timing before rendering. A transcript from different audio is not applied automatically.

### How many credits does Speech to Text use?

Speech to Text draws 4 credits per run.

_Last updated October 15, 2026_