Ollama / Ollama Cloud → Cherry Studio
Follow the confirmed official links and verified configuration step by step. This site never creates keys or sends model inference calls for you.
Step 1: Prepare
Prepare the official account, your own key, or the local environment before entering configuration.
No personal API key required
Open the right setup entry in the application first
A menu path or command appears here only when official sources explicitly prove it. Open the verified place first, then enter the verified settings below.
Then select Ollama / Ollama Cloud
This in-app entry passed the public evidence gate; it was not guessed from configuration field names.Step 3: Enter each configuration field
Start with what each setting means, then copy or replace the value. These plain-language labels explain the role of the verified field; they do not claim the app uses the same menu wording.
These fields come from a verified official integration. Replace fields marked “Use your value”; enter the other verified values exactly as shown.
mode
Tells the tool which connection mode to use. Copy a verified fixed value exactly as shown.
local_ollama
This is a fixed value from the verified configuration. Enter it exactly as shown.
model
Tells the tool which model to use. When marked “Use your value,” replace it with a model identifier you actually have installed or can access.
<YOUR_OLLAMA_MODEL>
This is a placeholder. Replace it with your own key, model, address, or other matching value indicated by the field.
apiBase
The address where the tool sends requests. Copy a verified fixed value exactly; replace it only when it is shown as a placeholder.
http://localhost:11434
This is a fixed value from the verified configuration. Enter it exactly as shown.
provider
Tells the tool which AI provider or local service this connection uses.
ollama
This is a fixed value from the verified configuration. Enter it exactly as shown.
Show the full verified configuration template
{
"mode": "local_ollama",
"model": "<YOUR_OLLAMA_MODEL>",
"apiBase": "http://localhost:11434",
"provider": "ollama"
}
Step 4: Test that it works
Save the settings, then run one minimal test in your own application. This site does not send model inference calls for you.
- Save the settings and confirm the provider / model selected in the application is the one you just configured.
- Send a very short test message, for example asking the model to reply only with “OK”.
- If you receive a normal model reply with no authentication, quota, or connection error, the basic integration is working.
Step 5: Troubleshoot by error type
Start with the error type instead of randomly changing fields.
When a key is required, check that it is your own key created on the official page, that no extra spaces were copied, and that the account or project has permission. If this integration is marked no-key, make sure you did not select a cloud or key-required mode instead.
Return to the provider page and check the confirmed free allowance, reset period, and eligibility. When quota is exhausted or rate limited, wait for reset or choose another verified free option; do not assume free usage is unlimited.
Compare each verified field, protocol, and the official docs above. For a local setup, confirm the local service and model are running first; do not invent a base URL, port, or model name.