> ## Documentation Index
> Fetch the complete documentation index at: https://docs.stenox.app/llms.txt
> Use this file to discover all available pages before exploring further.

# File Transcription

> Transcribe audio and video files

## Transcribe Files

Stenox can transcribe audio and video files — not just live dictation. Perfect for meetings, interviews, podcasts, or any recorded content.

<Tip>
  File transcription uses your configured transcription provider. For best results with long files, use a cloud provider like DeepGram or Groq.
</Tip>

## Supported Formats

### Audio Files

| Format | Extension |
| ------ | --------- |
| MP3    | .mp3      |
| WAV    | .wav      |
| M4A    | .m4a      |
| AAC    | .aac      |
| FLAC   | .flac     |
| OGG    | .ogg      |
| WMA    | .wma      |
| AIFF   | .aiff     |
| CAF    | .caf      |

### Video Files

| Format | Extension |
| ------ | --------- |
| MP4    | .mp4      |
| MOV    | .mov      |
| AVI    | .avi      |
| M4V    | .m4v      |
| WebM   | .webm     |
| MKV    | .mkv      |

<Note>
  For video files, Stenox extracts and transcribes the audio track only.
</Note>

## How to Transcribe a File

<Steps>
  <Step title="Open Stenox">
    Click the Stenox menu bar icon
  </Step>

  <Step title="Select Transcribe File">
    Click **Transcribe File** in the dropdown menu
  </Step>

  <Step title="Choose Your File">
    Select an audio or video file from Finder
  </Step>

  <Step title="Wait for Processing">
    Stenox will transcribe and optionally enhance the text
  </Step>

  <Step title="Review Results">
    The transcription appears in the Stenox window. Copy or save as needed.
  </Step>
</Steps>

## Provider Considerations

### For Short Files (\< 5 minutes)

Any provider works well:

* **WhisperKit** — Local, private
* **Groq Whisper** — Fast, free tier
* **DeepGram** — Accurate

### For Long Files (> 5 minutes)

Cloud providers recommended:

* **DeepGram** — Best for long-form content
* **Groq Whisper** — Fast processing
* **AssemblyAI** — Good for meetings

<Warning>
  WhisperKit can handle long files but processing time increases significantly. For a 30-minute file, expect 10-15 minutes of processing time locally.
</Warning>

## Tips for Best Results

### Audio Quality

* Use the highest quality recording available
* Clear audio with minimal background noise transcribes best
* Multiple speakers work best with DeepGram or AssemblyAI

### File Size

* Smaller file sizes upload and process faster
* Consider compressing large files before transcription
* MP3 is a good balance of quality and size

### Language

* Set the correct language in your Profile settings
* Use multilingual models for non-English content
* Auto-detect works but explicit setting is more accurate

## AI Enhancement

File transcriptions can be enhanced just like live dictation:

* **Grammar correction** — Fix spoken language patterns
* **Punctuation** — Add proper punctuation
* **Formatting** — Clean up for readability

<Tip>
  For meeting transcripts, use a custom prompt like: "Format as meeting notes with bullet points for action items."
</Tip>

## Usage Limits

File transcription counts against your plan limits:

| Plan          | Limit                                 |
| ------------- | ------------------------------------- |
| Free          | 5 minutes/day total (including files) |
| Plus          | Unlimited                             |
| Lifetime      | Unlimited (with BYOK)                 |
| Lifetime Plus | Unlimited                             |

<Note>
  Using BYOK providers bypasses Stenox limits. Only your provider's limits apply.
</Note>

## Troubleshooting

<AccordionGroup>
  <Accordion title="Transcription is very slow" icon="hourglass">
    * Use a cloud provider for faster processing
    * Switch from WhisperKit to Groq or DeepGram
    * For local processing, use base.en instead of small.en
  </Accordion>

  <Accordion title="File format not supported" icon="file-circle-xmark">
    Convert your file to a supported format:

    * Use QuickTime (File → Export As) for video
    * Use online converters for audio
    * MP3 and MP4 are widely compatible
  </Accordion>

  <Accordion title="Poor transcription quality" icon="spell-check">
    * Check audio quality in the original file
    * Try a different provider (DeepGram often works best)
    * Set explicit language instead of auto-detect
    * Add technical terms to Vocabulary
  </Accordion>

  <Accordion title="File too large" icon="file-arrow-up">
    * Compress the file before transcribing
    * Split into smaller segments
    * Use cloud providers for large files
  </Accordion>
</AccordionGroup>

## Next Steps

<CardGroup cols={2}>
  <Card title="BYOK Setup" icon="key" href="/setup/byok-configuration">
    Configure cloud providers for file transcription.
  </Card>

  <Card title="Vocabulary" icon="book" href="/setup/vocabulary">
    Add terms for better accuracy.
  </Card>
</CardGroup>