Skip to main content

Playground

Based on the Serverless capabilities, AtlasCloud provides simple and easy-to-use serverless model services. Currently, the models we support are as follows. In the near future, we will not only add support for other multi-modal and multi-scenario model services, but also enable the function of registering serverless endpoints to the playground, helping users to use their own models more conveniently and quickly.

Supported Models​

Chat Models​

OrganizationModel NameAPI Model StringUnit Price (1M Tokens)
DeepSeekDeepSeek R1deepseek-ai/DeepSeek-R1$5.00
DeepSeekDeepSeek R1 Distill Llama 70Bdeepseek-ai/DeepSeek-R1-Distill-Llama-70B$1.50
MetaMeta Llama3 8B Instructmeta-llama/Meta-Llama-3-8B-Instruct$0.20
QwenQwen2.5 32B InstructQwen/Qwen2.5-32B-Instruct$1.00

Usage Introduction​

  1. Click on Chat to enter the Playground Chat page.

  2. You can choose the model on your own, select a suitable Prompt (if you're not satisfied, you can even create your own exclusive Prompt). You can also adjust the input parameters according to your needs. Of course, you can simply use the default settings. After that, you can embark on this wonderful conversation journey.

Example of API Usage​

Create an API Key for authentication​

  1. Click Settings. After entering the account settings page, expand the drop - down box for API key management.

  2. Click the "Create API Key" button to create your own API key, as shown in the figure below.

Use third - party tools to call the Playground API​

Here, we use the desktop app "Chatbox" for demonstration. You can choose a suitable tool according to your preferences, such as cURL, OpenWebUI, or Cherry Studio.

  1. After downloading and installing the Chatbox app, open it and then open the settings page.

  2. Select the model provider as "Add Custom Provider", and then configure it according to the figure below.

Notes:

  • The API Host should be https://api.atlascloud.ai/v1. The /v1 base URL must be included.
  • The API Key is mandatory, and its value should be obtained from the previous step.

  1. Use the chat function of Chatbox, as shown in the figure below.

Billing Rules​

  1. The official models are priced based on millions of tokens.
  2. The system counts the number of tokens used by the model once every hour and deducts the fees accordingly.
  3. Please refer to the billing document for detailed billing rules.