> For the complete documentation index, see [llms.txt](https://speaking-test-docs.speechace.com/speechace-speaking-test/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://speaking-test-docs.speechace.com/speechace-speaking-test/speechace-workspaces-create-manage-and-share-speaking-assessments/creating-assessments/custom-assessments/roleplay-assessments/roleplay-settings.md).

# Roleplay Settings

The platform will then automatically generate the below settings based on your roleplay description. It is recommend for the test creators to review a few key settings carefully to make sure they align with the experience they want to create.

Based on the roleplay scenario description, the app automatically selects an appropriate interviewer role for the AI and an interviewee role for the test taker. Test creators have the option to modify the role as per their needs.

## AI Interviewer Mode 

AI Interviewer Mode allows test creators during an assessment to configure how a candidate interacts with the AI interviewer. Candidates can interact with the assessment using one of three modes: **Voice + Text, Voice, or Text**. Depending on the assessment requirements, the test creator can select the interaction mode that best fits the assessment requirement .

***

### Voice + Text

In **Voice + Text** mode, candidates can hear the AI interviewer’s questions through audio while also viewing the question text on the screen.

![](https://2079371467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKCjGnWORTEi39HEr4hwY%2Fuploads%2FC5hYwhSf5gtBAsmheGi1%2Funknown.png?alt=media\&token=99176b0b-7249-40ee-8e6f-158e6ddc2c1d)

This mode makes it easier for the candidate to understand the AI interviewer’s questions clearly. It also supports a more accessible and flexible assessment experience by allowing candidates to listen to and read each question before responding.

![](https://2079371467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKCjGnWORTEi39HEr4hwY%2Fuploads%2F69fZdUBf5gtSoSM2LZ8l%2Funknown.png?alt=media\&token=46fe27ef-3b96-4482-a749-1e306c59e736)

***

### Voice

In **Voice** mode, candidates interact with the AI interviewer through audio. The interviewer’s questions are played as voice without displaying the corresponding question text.<br>

![](https://2079371467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKCjGnWORTEi39HEr4hwY%2Fuploads%2F6cQrZwTW8nYYDJOZtyZ0%2Funknown.png?alt=media\&token=1ffa6f7d-01c0-42f6-a055-f37cdd7f6f07)

This mode supports assessment integrity by presenting the AI interviewer’s questions only through audio, thereby preventing candidates from reading or capturing the question text. It also encourages candidates to listen carefully and respond in real time.

![](https://2079371467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKCjGnWORTEi39HEr4hwY%2Fuploads%2FNcuhyEAfnqU9U2KzZdA6%2Funknown.png?alt=media\&token=b31864da-94da-413f-842d-6ecb245ac78b)

***

### Text

In **Text** mode, the AI interviewer’s questions are displayed as text on the screen without voice playback.

![](https://2079371467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKCjGnWORTEi39HEr4hwY%2Fuploads%2FCkAL128X0haoKrQUdtT3%2Funknown.png?alt=media\&token=dcf9463a-dc6d-48f6-8c50-abdd83f7fe07)

This mode allows candidate to interact with the assessment entirely through text. It is particularly useful for evaluating written communication and comprehension skills while providing a consistent experience to the candidate.

![](https://2079371467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKCjGnWORTEi39HEr4hwY%2Fuploads%2FYCwxcIYc0Jp87YUmRf33%2Funknown.png?alt=media\&token=1d6849b8-9292-4cff-ac39-e581f2b7bd44)

***

## Skills&#x20;

Based on the roleplay description, the app automatically selects a certain set of skills to assess. Test creators have the option to modify the skills as per their needs.

## Roleplay Input Type

Using this settings, test creators can specify if the test takers can respond to the roleplay conversation using one of voice, text or both.

Once the settings have been configured, test creators can now configure the structure or interactions of the roleplay as presented in the next section.