Cherry Studio
Cherry Studio is a desktop AI client. Start with OpenAI-compatible chat, then enable additional features only when your selected model supports them.
Prerequisites
Install Cherry Studio and prepare a Fenno API key plus a model ID that supports Chat Completions. Copy the ID from the Fenno console or model list.
1. Add a provider
Open Settings → Model Service → Add. Name the provider Fenno and select OpenAI, or the equivalent OpenAI-compatible type in your version.
| Setting | Value |
|---|---|
| Provider name | Fenno |
| API key | Your Fenno API key |
| API address | https://api.fenno.ai/v1 |
| Model ID | Exact ID copied from the console |
Keep the default Chat Completions mode for the first test. Do not enable several protocols or extra parameters at once.
2. Add and enable models
- Open Manage or Fetch models in the provider settings.
- Add the models you want to the enabled list. If discovery fails, add an exact ID manually.
- Enable the provider, return to chat and select a model under Fenno.
An entry in the model picker confirms configuration, not permission to call that model with your current key.
3. Test a conversation
Run the client’s connection check with your chosen model. Then start a new conversation, send “Reply with OK only”, and verify the corresponding request in the Fenno console.
If discovery fails but a manually added model works, check which model and protocol the discovery test uses. A provider icon alone is not proof of connectivity.
Responses and other capabilities
If your version offers Responses mode, enable it only for a model that supports Responses. The base URL remains https://api.fenno.ai/v1.
Image input, image generation and knowledge-base embeddings are separate capabilities. They may use different models and endpoints. Configure and test them separately after chat works.
Troubleshooting
| Symptom | Check |
|---|---|
| Provider missing from the picker | Enable the provider and add models to its enabled list |
| Empty model list | Check the URL, key and access; try an exact ID manually |
| 404 / missing route | The chat request must target /v1/chat/completions, without a repeated /v1 |
| Chat works but images or tools fail | Check model capabilities, protocol and client feature settings |
| 401 / 429 | See errors and retries rather than repeatedly running connection tests |
Menu labels vary by version. See the official Cherry Studio documentation .
