Gemini Setup
Get API Key
- Visit Google AI Studio
- Log in with your Google account
- Click "Get API Key"
- Create a new API Key
Add Channel in CCX
Gemini Endpoint (Native Protocol)
| Field | Value |
|---|---|
| Name | Gemini |
| Service Type | gemini |
| Base URL | https://generativelanguage.googleapis.com |
| API Keys | Your Google AI API Key |
Steps
- Go to CCX admin console, select Gemini endpoint
- Click "Add Channel"
- Set service type to
Gemini - Set Base URL to
https://generativelanguage.googleapis.com - Paste your API Key
- Save
Chat Endpoint (OpenAI Compatible)
Gemini can also be accessed via OpenAI-compatible interface:
| Field | Value |
|---|---|
| Name | Gemini (OpenAI) |
| Service Type | openai |
| Base URL | https://generativelanguage.googleapis.com/v1beta/openai |
| API Keys | Your Google AI API Key |
Available Models
| Model | Description |
|---|---|
gemini-2.5-pro | Latest flagship with thinking |
gemini-2.5-flash | Fast version with thinking |
gemini-2.0-flash | High-speed multimodal |
gemini-1.5-pro | Ultra-long context (2M tokens) |
Advanced Configuration
thought_signature Handling
Some third-party Gemini proxies require thought_signature in requests:
- Inject dummy thought_signature: Enable
injectDummyThoughtSignature - Strip thought_signature: Enable
stripThoughtSignature(for older APIs)
Notes
- Gemini native endpoint uses Google's generateContent protocol
- For OpenAI Chat protocol access, use Chat endpoint +
openaiservice type + OpenAI-compatible Base URL - Servers in mainland China need a proxy to access Google APIs
- Vision capabilities are automatically supported via native protocol