HopBase DocsThird-party clients
Configure HopBase in Cherry Studio, NextChat, LobeChat, and similar clients.
Desktop and web clients such as Cherry Studio, NextChat, and LobeChat use the same core settings. The provider type must match the model protocol.
| Model | Provider type | API address / Base URL | API Key |
|---|---|---|---|
| GPT, GLM, and Gemini models such as gpt-5.6-sol, glm-5.3, or gemini-3.8-flash | OpenAI (Chat Completions) | https://api.hop-base.com/v1 | Key from the matching group |
| Claude models such as claude-sonnet-5 | Anthropic | https://api.hop-base.com without /v1 | Claude Max (Official, Full Quota) or Claude Max (Official, Full Quota) - Limited-Time Offer group key; Claude Max (ccmax) works only in Claude Code |
Configuration steps
Confirm an available model
Call GET /v1/models with the target plan key and copy the complete model ID.
Add a custom provider
Enter the provider type, Base URL, API key, and model ID from the table above.
Save and test
Select the new model and send “Reply only with ok.” For a 404, check whether Claude was configured as OpenAI or whether /v1 is missing or duplicated.
Do not select OpenAI for a Claude plan key. It returns 404 because the request uses the wrong protocol path. Select Anthropic, use a client that supports Anthropic such as Cherry Studio, or use the web console.
Direct browser connections are supported, but web clients often do not append an API path. For OpenAI, GPT, GLM, or Gemini mode, enter the complete https://api.hop-base.com/v1 URL. For Anthropic mode and Claude, omit /v1.
Next steps
- Editors and IDEs: configure Cursor, Cline, or Continue
- Available models: find the model IDs available to a key
- Troubleshooting: look up an error message