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

# Real-time Transcription

> Real-time transcription specifications using the Vatis Streams API

### <Icon icon="sliders" iconType="light" size="24" />  Stream configuration template ids

* production: `670ba9e0efa59fe6aecd56f1`

### <Icon icon="upload" iconType="light" size="24" />  Input data

* multimedia content: `wav`, `mp3`, `mp4`, `webm`, `flv`, etc.

### <Icon icon="download" iconType="light" size="24" />  Egress responses

* [Egress Response](/api-reference/schemas/egress/stream-sink-message)
  * [Transcription Response](/api-reference/schemas/egress/transcription-response)
  * [LLM Response](/api-reference/schemas/egress/llm-response)

### <Icon icon="file-code" iconType="light" size="24" />  Export response

* [Audio Intelligence Response](/api-reference/schemas/export/audio-intelligence-response)

### <Icon icon="wrench" iconType="light" size="24" />  Parameters

<AccordionGroup>
  <Accordion title="Transcription">
    <ParamField query="language" type="string" default="en">
      The language as country [ISO code](/speech-to-text/real-time/transcription-languages) to transcribe the audio to. When not specified, it is automatically detected.
    </ParamField>

    <ParamField query="multichannel" type="bool" default="false">
      Enable or disable multichannel audio processing. Possible values: `true` or `false`.
      When enabled, each channel is processed separately.
    </ParamField>

    <ParamField query="utteranceEnd" type="int" default="1000">
      The minimum waiting time in milliseconds after the last speech is detected before the emitting an `utterance` transcription message.
    </ParamField>

    <Accordion title="Raw data format parameters">
      <Warning>Use these parameters only when sending raw data.</Warning>

      <ParamField query="encoding" type="string" default="none">
        The encoding of the audio data. Accepted values:

        * `linear16`: 16-bit signed little-endian samples (Linear PCM)
        * `flac`: Free Lossless Audio Codec
        * `mulaw`: 8-bit samples (G.711 mu-law)
        * `opus`: OGG Opus
      </ParamField>

      <ParamField query="sampleRate" type="int" default="none">
        The sample rate of the audio data in Hertz (Hz).
      </ParamField>

      <ParamField query="channels" type="int" default="none">
        The number of audio channels in the audio data.
      </ParamField>
    </Accordion>
  </Accordion>

  <Accordion title="Audio Intelligence">
    <ParamField query="summary" type="bool" default="false">
      Create a summary based on the upstream content.
    </ParamField>

    <ParamField query="summaryLength" type="string" default="brief">
      Summary length. Options are: `brief` or `detailed`.
    </ParamField>

    <ParamField query="summaryTone" type="string" default="conversational">
      Summary tone. Options are: `conversational` or `informative`.
    </ParamField>

    <ParamField query="summaryStructure" type="string" default="paragraphs">
      Summary structure. Options are: `paragraphs` or `bullet_points`.
    </ParamField>

    <ParamField query="sentimentAnalysis" type="bool" default="false">
      Perform sentiment analysis on the upstream content.
    </ParamField>

    <ParamField query="ask[0-N]" type="string" default="none">
      Specifies the custom prompt content for one of the `ask0`, `ask1`, ..., `askN` ask anything slots.
      When the content is specified, the prompt is considered activated. Otherwise, it is deactivated.
    </ParamField>

    <ParamField query="ask[0-N]System" type="string" default="none">
      Specifies the custom system prompt for one of the `ask0System`, `ask1System`, ..., `askNSystem` ask anything slots.
    </ParamField>

    <ParamField query="ask[0-N]Id" type="string" default="none">
      Specifies the prompt id for one of the `ask0Id`, `ask1Id`, ..., `askNId` ask anything slots.
      The role of the prompt id is to identify the prompt in the responses.
      When unspecified, it will fallback on the index of the slot (e.g. `"0"`, `"1"`).
    </ParamField>

    <ParamField query="ask[0-N]Format" type="string" default="none">
      Specifies the prompt response [JSON Schema](https://json-schema.org/overview/what-is-jsonschema) encoded as a string for one of the `ask0Format`, `ask1Format`, ..., `askNFormat` ask anything slots.

      When unspecified, the response not be structured in any particular way.
      When specified, the response will be a JSON object encoded as a string.
    </ParamField>

    <ParamField query="enhancedTranscription" type="bool" default="false">
      Enable transcription enhancement.
    </ParamField>

    <ParamField query="etModel" type="[standard|enhanced]" default="standard">
      Balance the speed-accuracy ration of the transcription enhancing model.
    </ParamField>

    <ParamField query="etVocabulary" type="string[]" default="null">
      Provide a list of words to enhance the transcription. Those can be domain-specific terms that the model should pay special attention to.
    </ParamField>

    <ParamField query="etSystemPrompt" type="string" default="null">
      Provide a system prompt to guide the transcription enhancement model. This can help the model understand the context or specific requirements for the transcription.
    </ParamField>
  </Accordion>
</AccordionGroup>