How to use Cherry Studio with a custom OpenAI or Anthropic API host
Add a custom provider in Cherry Studio with OpenAI endpoint https://www.ccsub.net/api/v1 and Anthropic endpoint https://www.ccsub.net, and how the /v1 and # rules work.
Copy-paste configuration
These are the three values the tool needs. Create the API key on the API Keys page of your CCSub dashboard.
- OpenAI endpoint
https://www.ccsub.net/api/v1（already has /v1, nothing appended）- Anthropic endpoint
https://www.ccsub.net（becomes /v1/messages）- API Key
sk-your-ccsub-key- Model
claude-sonnet-5（use Sync models to fetch the list）
For model, use an ID such as claude-sonnet-5 — the full list of model IDs is on the models page. No key yet? Create an account first.
Steps at a glance
- 1
Create a CCSub API key
Sign up for CCSub and create an sk- key on the API Keys page of the dashboard.
- 2
Add a provider
In Cherry Studio, open Settings → Model Provider and click Add Provider. In Add Custom Provider, name it CCSub and paste the key.
- 3
Fill in the endpoints
Under Endpoint settings, set OpenAI to https://www.ccsub.net/api/v1 and Anthropic to https://www.ccsub.net, and check the Request path preview under each field.
- 4
Sync models and check
On the provider page, click Sync models to load the model list, then Check to test the connection.
- 5
Start a chat
Open a new chat, pick a model under CCSub such as claude-sonnet-5, send a message, and confirm it on the CCSub Usage page.
Cherry Studio is an open-source desktop AI client for Windows, macOS and Linux. In version 2.x, a single provider can carry both an OpenAI endpoint and an Anthropic endpoint under one API key. The one thing to understand before connecting it to CCSub is that it appends /v1 to the address you type.
Add CCSub as a provider (2.x)
- Open Settings → Model Provider and click Add Provider.
- Fill in the Add Custom Provider panel:
- Provider name
CCSub- API Key
sk-your-ccsub-key- Endpoint settings → OpenAI
https://www.ccsub.net/api/v1Preview should read /api/v1/chat/completions.- Endpoint settings → Anthropic
https://www.ccsub.netPreview should read /v1/messages.
- Each endpoint field shows a Request path: preview underneath. Compare it with the table above. The first endpoint you fill in is marked Default; use Set as default to change that.
- Save, then on the provider page click Sync models to load CCSub's model list, and Check to test the connection.
The More options section holds OpenAI Responses, Gemini and image endpoints. You do not need them for chatting through CCSub.
How the /v1 auto-append works
This is where most Cherry Studio setups go wrong, and the rule has changed over time:
- Default
appends /v1A trailing / is removed first, then /v1 is added.- Path already has a version
left alonee.g. /api/v1 or /v1 stays as typed.- Ends with #
left aloneUI hint: Add # at the end to disable the automatically appended API version.
So https://www.ccsub.net/api/v1 already contains /v1 and is not changed, while https://www.ccsub.net gets /v1 added. Both end up at a working CCSub endpoint.
"Trailing slash skips v1" is the old rule. Cherry Studio 1.6 and earlier told you that ending the address with / would skip the /v1, and plenty of tutorials still say so. Today the slash is simply stripped and /v1 is still added. An old-style https://www.ccsub.net/v1/ therefore still resolves to /v1/chat/completions, which is fine — but anything like https://www.ccsub.net/v1/v1 returns a 404.
Cherry Studio 1.x
Older 1.x builds ask for a provider type when you add a provider. Choose OpenAI or Anthropic and use the same addresses as above. Models are fetched from the Manage dialog, and the API key, API host and Check controls work the same way as in 2.x.
Start chatting
Open a new conversation and pick a model listed under CCSub. The model field takes IDs such as claude-sonnet-5; the full list of IDs is on the models page. After your first message, the request and its charge appear on the Usage page of the CCSub dashboard.
If you also use terminal tools such as Claude Code or Codex, CC Switch can manage their provider settings in one place. To use CCSub inside VS Code or JetBrains, see the Continue guide.
Common errors and fixes
Check fails with 404
Cause：/v1 is doubled (e.g. https://www.ccsub.net/v1/v1) or /chat/completions was pasted into the host.
Fix：Stop the OpenAI endpoint at https://www.ccsub.net/api/v1 and confirm the preview reads /api/v1/chat/completions.
Sync models returns nothing or an HTML page
Cause：The /models path built from your address points at a website page instead of the API.
Fix：Use https://www.ccsub.net/api/v1 so the list is fetched from /api/v1/models.
401 authentication_error: Invalid or disabled API key
Cause：The key picked up whitespace when pasted, or it is disabled in the CCSub dashboard.
Fix：Paste the key again and make sure it is enabled.
402 insufficient_balance
Cause：Your balance is too low, or a monthly plan's daily quota is used up.
Fix：Top up; monthly quotas reset at 00:00 Beijing time (UTC+8).
For more error messages and what they mean, see the error troubleshooting index. Installation and network issues are covered in the troubleshooting docs (Chinese).
FAQ
Does a trailing / still skip the /v1?
No. That was the rule in Cherry Studio 1.6 and earlier. Current versions strip a trailing slash and then append /v1, unless the path already contains a version such as /v1 or the address ends with #. Many older tutorials still describe the slash rule.
Can I use https://www.ccsub.net as the OpenAI endpoint?
Yes — it becomes https://www.ccsub.net/v1/chat/completions, which works. Our docs use https://www.ccsub.net/api/v1; because its path already contains /v1, Cherry Studio leaves it alone.
What does a trailing # do?
It tells Cherry Studio to use the address as typed without adding a version. If you enter a full endpoint followed by # (for example .../chat/completions#), that exact URL is requested. You do not need it for CCSub.
How do I set this up in Cherry Studio 1.x?
In 1.x you choose OpenAI or Anthropic from a provider type dropdown and use the same addresses as above. Models are loaded through Manage, and the Check button works the same way.
Which model IDs can I use?
Sync models pulls every model ID from CCSub, or add IDs such as claude-sonnet-5 by hand. The full list is on the CCSub models page.
Related setup guides
Need a CCSub API key?
Sign up, create a key in the dashboard, and paste the values above into your tool. Pay-as-you-go balance does not expire — you only pay for the tokens you use.
