Skip to main content
Cherry Studio manages model providers separately from chat assistants. Add APIONE first, register the models, then select one in a conversation.

Prepare

Install Cherry Studio and create a key on the APIONE API Keys page. Start with a text model that supports Chat Completions.

Create the connection

  1. Open Settings → Model Provider and click Add.
  2. Set the name to APIONE and the type to OpenAI.
  3. Enter your APIONE key. Set API Host to https://api.apione.ai/v1/.
  4. Save, open Manager, refresh the model list, and select the models you need.
  5. Return to chat, choose an APIONE model, and send a simple question.
Add an APIONE provider in Cherry Studio and use the address for your selected protocol.

Add an APIONE provider in Cherry Studio and use the address for your selected protocol.

URL and model

Different protocols use the corresponding base address. OpenAI-compatible requests should ultimately reach https://api.apione.ai/v1/chat/completions; Anthropic requests should ultimately reach https://api.apione.ai/v1/messages. Check the path in the client address preview.

Use the Anthropic protocol

  1. Add an APIONE provider under Settings → Model Provider and choose Anthropic.
  2. Enter the APIONE key and use https://api.apione.ai as the API address.
  3. Confirm that the request preview is https://api.apione.ai/v1/messages, then enable the provider.
Highlighted APIONE key and API address in Cherry Studio

APIONE provider key and address configuration; the Anthropic address preview shows /v1/messages.

Click Fetch model list, search for the models you need, and add them. Model IDs and access permissions are based on the results returned by the current key.
Cherry Studio APIONE model search and selection dialog

After fetching the model list, search for and select APIONE models.

Return to the chat page, select a model under the APIONE provider at the top of the assistant, and send a short question. The image below shows a conversation with a selected model and a received reply.
Cherry Studio conversation using an APIONE model

Select an APIONE model in an assistant conversation and view the reply.

Model names in screenshots are for demonstrating the operation position; the actual selection should be based on models available to the current account. If automatic list fetching fails, copy the exact ID from the APIONE model directory and add it manually.

Verify before enabling features

Verify plain text before testing images, attachments, and tools. A listed model does not necessarily support every input type.
  • Authentication fails: check whitespace, key status, expiration, and model permissions.
  • Model not found: use the actual model ID in the model parameter, not its display name. A request ID identifies one API call and does not select a model.
  • Connection test passes but chat fails: confirm the assistant uses APIONE and disable unsupported optional parameters.
Integration directory · Choose a model · Troubleshooting