---
title: "Run and manage Playground tasks"
slug: "clovastudio-playground01"
updated: 2026-09-17T09:33:47Z
published: 2026-09-17T09:33:47Z
canonical: "guide.ncloud-docs.com/clovastudio-playground01"
stale: true
---

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

# Run and manage Playground tasks

Available in Classic and VPC

The Task Management section explains how to run and save tasks in **Playground** and how to view URLs for sharing saved tasks. It also explains how to use reasoning models that include thought processes and vision models that process images.

## Run a task

To run a task in Playground:

1. In the NAVER Cloud Platform console, navigate to **Menu** > **All Services** > **AI Services** > **CLOVA Studio**.
2. Click **My Product**, then click **[Go to CLOVA Studio]**.
3. Click **Playground**.
4. Configure parameters.
  - For details about parameters, see [Parameters](/docs/en/clovastudio-info#%ED%8C%8C%EB%9D%BC%EB%AF%B8%ED%84%B0).
  - In most cases, set Top P to 0.8-1.0 and Top K to 0.
  - Repetition Penalty prevents repetitive output. Values between 1.0-1.1 with adjustments of 0.05 are recommended.
5. Enter a task title.
6. In System Prompt, enter instructions or examples describing what the model should do.
7. Enter your message in the input field at the bottom of the interface.
              Note

              

Press **[Enter]** or **[Shift]**+**[Enter]** to insert a line break.
8. If needed, click **[Add turn]** to ask additional questions.
9. To check the result, click the ![playground_ execution.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/playground_%20execution.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D) or use the shortcut (**[Ctrl]**+**[Enter]**/**[⌘]**+**[Enter]**).
  - If you are not satisfied with the output, clear the content and run again to generate new results, or edit your examples and run again.
10. In the editor, you can switch between User and Assistant roles.
  - This is useful for quickstart creation or testing but is optional.
11. Depending on the action you want to perform, proceed with one of the following:
  - [Save task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EC%A0%80%EC%9E%A5)
  - [Share task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EA%B3%B5%EC%9C%A0)
  - [Load task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EB%B6%88%EB%9F%AC%EC%98%A4%EA%B8%B0)

### Use reasoning models

To use a reasoning model in Playground:

1. Click **Playground**.
2. In Model, select **HCX-007**.
  - HCX-007 supports text-only requests.  

![chatmode_thinking_ko.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/chatmode_thinking_en.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D)
3. Enter a task title.
4. In System Prompt, enter instructions or examples describing what the model should do.
5. Select a **Thinking** option.
  - This setting controls how long the model thinks before responding.
  - Options are **short, medium, and deep.** The default is **short**. If no thinking process is needed, select Off.  

![clovastudio-playground_think_ko.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/clovastudio-playground_think_en.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D)
6. Enter your message in the input field at the bottom.
              Note

              

Press **[Enter]** or **[Shift]**+**[Enter]** to insert a line break.
7. If needed, click **[Add turn]** for additional questions.
8. To check the result, click the ![playground_ execution.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/playground_%20execution.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D) or use the shortcut (**[Ctrl]**+**[Enter]**/**[⌘]**+**[Enter]**).
  - If you are not satisfied with the output, clear the content and run again to generate new results, or edit your examples and run again.
9. In the editor, you can switch between User and Assistant roles.
  - This is useful for quickstart creation or testing but is optional.
10. Depending on the action you want to perform, proceed with one of the following:
  - [Save task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EC%A0%80%EC%9E%A5)
  - [Share task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EA%B3%B5%EC%9C%A0)
  - [Load task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EB%B6%88%EB%9F%AC%EC%98%A4%EA%B8%B0)

              Caution

              

Playground displays only the final outputs. It does not show the model's internal reasoning steps. To view detailed reasoning, use the [Chat Completions V3 API](https://api.ncloud-docs.com/docs/en/clovastudio-chatcompletionsv3-thinking).

### Use vision models

To use a vision model in Playground:

1. Click **Playground**.
2. In **Model**, select **HCX-005**.
  - The upload button is enabled only when HCX-005 is selected.
  - If another model is selected, an error message is shown when uploading an image.
  - Images can be uploaded only in the user turn.  

![clovastudio_playground_vision_ko.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/clovastudio_playground_vision_en.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D)
3. Enter a task title.
4. In System Prompt, enter instructions or examples describing what the model should do.
5. Upload images using one of the following methods:
  - **File upload**: Select an image from your device
  - **Add link**: Enter an image URL  

![chatmode_vlm_upload_ko.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/chatmode_vlm_upload_en.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D)
  - Drag and drop: Drag an image into the input area  

![chatmode_vlm_upload2_ko.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/chatmode_vlm_upload2_en.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D)
6. Enter your message in the input field.
              Note

              

Press **[Enter]** or **[Shift]**+**[Enter]** to insert a line break.
7. If needed, click **[Add turn]** to ask additional questions.
8. To check the result, click the ![playground_ execution.png](https://cdn.document360.io/6998976f-9d95-4df8-b847-d375892b92c2/Images/Documentation/playground_%20execution.png?sv=2026-02-06&spr=https&st=2026-11-01T18%3A36%3A42Z&se=2026-11-01T19%3A36%3A42Z&sr=c&sp=r&sig=YH0UfPhDjWj3QYa%2BU7Nnp34A3PITwuQ9RqjIxst3row%3D) or use the shortcut (**[Ctrl]**+**[Enter]**/**[⌘]**+**[Enter]**).
  - If you are not satisfied with the output, clear the content and run again to generate new results, or edit your examples and run again.
9. In the editor, you can switch between User and Assistant roles.
  - This is useful for quickstart creation or testing but is optional.
10. Depending on the action you want to perform, proceed with one of the following:
  - [Save task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EC%A0%80%EC%9E%A5)
  - [Share task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EA%B3%B5%EC%9C%A0)
  - [Load task](/docs/en/clovastudio-playground01#%ED%94%8C%EB%A0%88%EC%9D%B4%EA%B7%B8%EB%9D%BC%EC%9A%B4%EB%93%9C%EC%9E%91%EC%97%85%EB%B6%88%EB%9F%AC%EC%98%A4%EA%B8%B0)

              Caution

              

Playground does not support saving tasks that contain images.

#### Conditions for uploading images

The upload requirements for images are as follows:

| Item | Description |
| --- | --- |
| Number of uploads | - One image per turn, - up to five images per conversation |
| Formats | - BMP (.bmp) - PNG (.png) - JPG (.jpg) - JPEG (.jpeg) - WEBP (.webp) |
| Size | More than 0 bytes, up to 20 MB |
| Aspect ratio | Up to 1:5 or 5:1 |
| Dimensions | - Longest side up to 2,240 px; shortest side at least 4 px - Images are automatically resized when uploaded |

              Caution

              

- You are solely responsible for the copyright and management of uploaded images. Ensure that you do not violate others' rights.
- Do not upload files containing personal information or illegal/inappropriate content.

## Save a Playground task

To save a task:

1. In CLOVA Studio, click Playground.
2. After writing your prompt, click **[Save]** or **[Save as]** in the upper-right corner.
  - **[Save]**: Save is used when saving a task for the first time or overwriting an existing saved task using the same name.
  - **[Save as]**: Save as is used to save the current task under a new name or to create a separate copy.
3. Enter a task name.
4. If needed, add tags.
  - To add tags, click **[Tags]**. The tag list opens and you can select tags related to your task.
  - Tags help with task search and classification.
5. You can view saved tasks from **My Tasks** on the left side of the CLOVA Studio interface.

## Load a Playground task

You can load sample tasks or your saved tasks in Playground to run them again or share them.

To load a task in Playground:

1. On the left side of CLOVA Studio, click **Playground**.
2. Click **[Load]**, then select the task you want to open.
  - Only samples or tasks you have saved can be loaded.
3. Adjust parameters or edit the title and prompt, then run the task again.

To load a task from My Tasks:

1. On the left side of CLOVA Studio, click **My Tasks**.
2. Under the **[Playground]** tab, click the title of the task you want to open.
3. If needed, enter a search term to find the task to open.
  - You are taken directly to the Playground interface with the selected task.

To load a task from Prompt Examples:

1. On the left side of CLOVA Studio, click Prompt Examples.
2. Click the task you want to open in Playground.
3. If needed, select tags or enter a search term to find the task.
  - You are taken directly to the Playground interface with the selected task.

## Share a Playground task

You can share tasks created in Playground with other users. To share a task, first save it, then click **[Share]**.

- You can view the share URL in the share popup window.
- A password required to access the shared task is also displayed.
- Click **[Copy]** to copy both the URL and the password to your clipboard.