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

# Dashboard

> Faheem landing page inside the dashboard. Hero with feature specs, How It Works strip, three transcription methods, supported formats, recent files, pro tips.

<Note>
  The Speech to Text dashboard lives at `/{lang}/speech-to-text`. It is a single page with five sections: hero, How It Works, Start Transcribing method cards, supported audio formats, and Pro Tips for best results.
</Note>

## Feature specs

<CardGroup cols={2}>
  <Card title="100+ Languages & Dialects" icon="globe">
    Arabic, English, French, Spanish, and 25+ Arabic dialects.
  </Card>

  <Card title="Real-Time Processing" icon="bolt">
    Sub-300ms latency for live transcription use cases.
  </Card>

  <Card title="Diarization & Gender Detection" icon="users">
    Identify speakers, separate voices, and detect gender.
  </Card>

  <Card title="Code-Switching Ready" icon="language">
    Arabic-English mixed speech handled natively.
  </Card>
</CardGroup>

## How It Works

A compact stepper strip. Five steps, each with a colored icon tile.

<Steps>
  <Step title="Upload or Record">
    Bring audio in by uploading a file, recording from the mic, or speaking live.
  </Step>

  <Step title="Detect Language">
    Faheem auto-detects the language. You can also override it from the language selector.
  </Step>

  <Step title="Faheem AI Engine">
    The Arabic-first ASR engine transcribes the audio with sub-300ms latency.
  </Step>

  <Step title="Review & Edit">
    Open the transcript editor, click any word to fix it, rename speakers, adjust cue times.
  </Step>

  <Step title="Export & Share">
    Quick-export to PDF, DOCX, or TXT, or copy the plain text.
  </Step>
</Steps>

## Start Transcribing

Three method cards in the **Start Transcribing** grid. Each opens a modal.

<CardGroup cols={3}>
  <Card title="Upload File" icon="cloud-arrow-up" href="/en/Systems/Speech-to-Text/Upload-Audio">
    **Drag & drop or browse audio files** , supports MP3, WAV, M4A, AAC, OGG, FLAC and more.

    Highlights: Drag & drop, Batch processing, All formats. Badge: **No setup required**. Button: **+ Upload Audio File**.
  </Card>

  <Card title="Live Transcription" icon="microphone-lines" href="/en/Systems/Speech-to-Text/Live-Transcribe">
    **Speak directly and watch your words appear in real-time.** Perfect for meetings, interviews, and notes.

    Highlights: Sub-300ms latency, Copy transcript, Language detection. Badge: **Instant**. Button: **+ Start Live Transcription**.
  </Card>

  <Card title="Record Audio" icon="microphone" href="/en/Systems/Speech-to-Text/Live-Transcribe">
    **Record from your microphone, review the audio, and transcribe it,** all in one seamless flow.

    Highlights: Record + transcribe, Audio playback, Auto-save. Badge: **No setup required**. Button: **+ Start Recording**.
  </Card>
</CardGroup>

<Tip>
  **Upload File** and **Record Audio** both open the **Live Transcribe** modal in record mode. They share the same backing flow.
</Tip>

## Supported Audio Formats

Strip card under the method cards. **All popular and professional audio formats supported.**

| Format | Extensions |
| ------ | ---------- |
| MP3    | `.mp3`     |
| WAV    | `.wav`     |
| M4A    | `.m4a`     |
| AAC    | `.aac`     |
| OGG    | `.ogg`     |
| FLAC   | `.flac`    |
| WEBM   | `.webm`    |

The dashboard also accepts MP4 audio (`.mp4`) when you go through Upload File.

## Pro Tips for best results

Seven numbered tips, two-column grid, each with a cyan badge.

<CardGroup cols={2}>
  <Card title="1. Verify detected language" icon="circle-check">
    Ensure the correct language is selected for better accuracy.
  </Card>

  <Card title="2. Don't upload sensitive content" icon="shield">
    Be mindful of shared or confidential information.
  </Card>

  <Card title="3. Review audio before submitting" icon="headphones">
    Check for clear audio quality and minimal background noise.
  </Card>

  <Card title="4. Wait for processing to complete" icon="hourglass">
    Larger files may take a few moments to transcribe.
  </Card>

  <Card title="5. Review and edit transcribed text" icon="pen-to-square">
    Manually correct any AI inaccuracies for perfect results.
  </Card>

  <Card title="6. Use supported formats (MP3, MP4 or WAV)" icon="file-audio">
    For best compatibility, use high-quality audio formats.
  </Card>

  <Card title="7. Avoid background noise & speak clearly" icon="microphone">
    Clearer input leads to significantly better transcription output.
  </Card>
</CardGroup>

## Sidebar swap

When you are on `/speech-to-text/*`, the dashboard shell swaps:

* **System Switcher** auto-syncs to **Speech to Text (Faheem)**.
* The sidebar shows **Your Files**, the live STT file list.
* **+ Create Agent** in the topbar becomes **+ Transcribe Audio**, with three options: Upload File (cyan), Live Transcription (green), Record Audio (red).
* The breadcrumb on the editor page reads **Speech to Text (Faheem)** › *file name*.

## Recent files

Below the formats strip the dashboard also surfaces a **Latest from the library** section pulled from the same source as the sidebar. Each row clicks through to the [Transcript Editor](/en/Systems/Speech-to-Text/Transcript-Editor) and exposes the [3-dot actions menu](/en/Systems/Speech-to-Text/File-Library) (export, pin, rename, delete).

## Common questions

<AccordionGroup>
  <Accordion title="Which method should I pick?">
    **Upload File** is for audio you already have on disk. **Live Transcription** is for transcribing while you speak (your browser does the work). **Record Audio** opens the same modal as Live Transcription, the difference is intent: Live Transcription is meant for note-taking, Record Audio is meant for capturing a session you'll review later.
  </Accordion>

  <Accordion title="Is Live Transcription the same quality as Upload File?">
    Upload File uses the Faheem engine and is generally more accurate, especially for Arabic and dialects. Live Transcription uses your browser's built-in speech recognition which varies by browser. For the most accurate transcript, record audio first then upload it.
  </Accordion>

  <Accordion title="What's the maximum length I can transcribe?">
    5 hours per file, with a 500 MB file size limit. If your audio is longer, split it into parts before uploading.
  </Accordion>

  <Accordion title="Where do my transcribed files go?">
    Into your [File Library](/en/Systems/Speech-to-Text/File-Library) and into the **Your Files** sidebar list (which only shows on Speech to Text pages). Each file is private to your Wittify account.
  </Accordion>

  <Accordion title="Why is Faheem's accuracy better for Arabic?">
    The engine was trained Arabic-first, with strong support for the major Arabic dialects (Egyptian, Saudi, Levantine, Maghrebi, Gulf, and others). It also handles code-switching between Arabic and English natively, which most generic engines don't do well.
  </Accordion>

  <Accordion title="Can I transcribe a meeting in two languages at once?">
    Yes, Faheem handles code-switched audio without you having to do anything special. Pick **Auto Detect** in the language dropdown, or pick the dominant language. The engine still picks up the second language wherever it appears.
  </Accordion>
</AccordionGroup>