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

# STT Playground

> Record live or upload audio to transcribe across providers and models

The STT Playground makes it easy to try out Inworld's Speech-to-Text capabilities through an interactive playground. Use it to compare transcription models, test language hints and custom vocabulary, and watch Voice Profile analyze the speaker in real time — all before writing a line of code.

<img src="https://mintcdn.com/inworldaidev/WP4-KQ5wo6zwAfSm/img/portal/stt-playground.png?fit=max&auto=format&n=WP4-KQ5wo6zwAfSm&q=85&s=aa09e919363e7c6834beb43e70e0225d" alt="STT Playground" width="2322" height="2004" data-path="img/portal/stt-playground.png" />

## Get Started

<Steps titleSize="h3">
  <Step title="Go to Inworld Portal" iconType="light">
    In [Portal](https://platform.inworld.ai/), select **Speech-to-Text** from the left-hand side panel.
  </Step>

  <Step title="Choose a model">
    Click the model dropdown at the top of the playground to browse available models. Each entry shows its provider and language coverage — from Inworld's own `inworld/inworld-stt-1` to models from Groq, AssemblyAI, Soniox, and Deepgram. See the [model comparison](/stt/overview#model-comparison) for details on each.
  </Step>

  <Step title="Record or upload audio">
    Press the mic button to transcribe live from your microphone, or use the upload button to transcribe an audio file. The status bar shows **Ready** when the playground can accept audio and **Listening…** while recording.

    Input options depend on the selected model's endpoint support — for example, `groq/whisper-large-v3` transcribes uploaded files only (no live streaming). See the [model comparison](/stt/overview#model-comparison) for each model's endpoints.
  </Step>

  <Step title="Review the transcript">
    The transcript appears in real time as you speak. When you're done, use the copy button to grab the text, or the code button to view the equivalent API request.
  </Step>
</Steps>

## Advanced Features

For greater control over transcription, try the following:

1. **Language hint** — The **Language** dropdown defaults to **Auto-detect**, which recognizes the spoken language automatically. Select a specific language to give the model a hint when you know the audio's language in advance — this generally improves accuracy, especially for short utterances. See [Language Support](/stt/languages) for how hints behave.
2. **Custom vocabulary** — Click **Vocabulary** to add key terms (names, jargon, acronyms) that bias recognition toward them. Terms are added as chips, and the button shows how many are active. This is a soft bias rather than a hard keyword lock — test it on the terms the baseline actually misses. The equivalent API field is [`prompts`](/stt/quickstart#make-your-first-stt-api-request).
3. **Voice Profile** — The panel on the right analyzes the speaker alongside transcription, showing **Age**, **Accent**, **Voice Pitch**, **Vocal Style**, and **Emotion**, each with a confidence score. It populates after transcription begins. See [Voice Profiles](/stt/voice-profiles) for the full category reference.

## From Playground to Code

Click **Integrate** (top right) — or **Get integration code** in the Voice Profile panel — to generate an API request matching your current playground configuration: model, language hint, vocabulary, and Voice Profile settings.

## Next Steps

Ready to build? Here's where to go next.

<CardGroup cols={3}>
  <Card title="Quickstart" icon="bolt" href="/stt/quickstart">
    Make your first STT API call in minutes.
  </Card>

  <Card title="Voice Profiles" icon="user" href="/stt/voice-profiles">
    Analyze speaker characteristics alongside transcription.
  </Card>

  <Card title="Language Support" icon="globe" href="/stt/languages">
    Language coverage and hinting behavior.
  </Card>
</CardGroup>