Cherry Studio
Last updated August 26, 2026
Add an OpenAI-type provider in Cherry Studio pointing at Amux; the model list fills itself in.
Cherry Studio is a desktop client for working with many models. Add an OpenAI-type provider pointing at Amux and it fetches the model list automatically, so there is no need to enter model IDs manually.
1. Create an Amux key
Create one on the keys page in the console. The key remains viewable there after creation, so you do not need to rotate it just to retrieve it again.
2. Add the provider
Go to Settings → Model Provider and click "Add":
| Field | Value |
|---|---|
| Provider name | Amux (anything — it is just the label in the list) |
| Provider type | OpenAI |
| API key | Your Amux key |
| API host | https://gateway.amux.ai/v1 |
Watch for leading or trailing whitespace when pasting the key — it is the most common way this step fails, and the error you get back says "invalid key", which looks like a problem with the key itself.
3. Fetch the model list
Click "Manage" once the provider is saved and Cherry Studio fetches the available models. It requests Amux's /v1/models, so the result includes every model available to that key, along with each provider's pricing and context length.
4. Check it works
Pick a model from the list, start a new conversation, and send a message.
Choosing models
Each entry is named by its canonical ID (anthropic/claude-opus-5 and so on) and is not limited to OpenAI's models. Claude, Gemini, and other compatible models can also be called over the OpenAI protocol. For parameter differences in cross-protocol calls, see protocol conversion compatibility.
A model often has several providers at different prices. Cherry Studio does not distinguish between them, and routing picks one according to the strategy in your settings. To pin one, add a model ID with the suffix by hand (anthropic/claude-opus-5:anthropic) — see the API reference.
Troubleshooting
The model list stays empty
Check the API host field first. Cherry Studio appends a path to whatever you enter, so the wrong shape returns nothing — it needs https://gateway.amux.ai/v1, with the /v1. Then check the key and your network.
"Invalid key"
Usually whitespace picked up while copying. Copy it again and check both ends after pasting. Also confirm the key is still enabled in the console and the workspace has balance.
One model fails while others work
Check the workspace balance, and whether that model is still listed — the list is a snapshot from when you fetched it and does not update itself when a model is retired. Click "Manage" to fetch it again.
Separating spend per project
Create a separate key per use and add it as a second provider in Cherry Studio. Usage is counted per key, and requests under one key cannot be split apart afterwards.