Kimi
Kimi is an AI assistant by Moonshot AI, known for its long context processing capabilities.
Supported Models
| Model | Description |
|---|---|
| Kimi-K2.5 | Latest Kimi K2.5 model |
Features
- Long context processing (up to 200K tokens)
- Web search integration
- Streaming/non-streaming output
- Multi-turn conversation support
- Tool calling support
- Free accounts do not support deep thinking (official limitation)
Authentication
Kimi supports OAuth Login, Refresh Token, and JWT Token authentication methods.
Method 1: OAuth Login (Recommended)
In Chat2API, click Add Provider and select Kimi.
Choose OAuth Login method.
Click Start Login, the system will automatically open the browser and redirect to the Kimi login page.
After completing login in the browser, return to Chat2API, the system will automatically obtain authentication information.
Method 2: Refresh Token
Go to Kimi and log in.
Press F12 to open Developer Tools.
Go to Application > Local Storage.
Find refresh_token and copy its value.
If refresh_token is an array, join the values with . (dot).
Method 3: JWT Token (kimi-auth Cookie)
Go to Kimi and log in.
Press F12 to open Developer Tools.
Go to Application > Cookies > https://www.kimi.com.
Find kimi-auth cookie and copy its JWT token value (starts with eyJ).
Adding to Chat2API
- Navigate to Providers > Add Provider
- Select Kimi from the list
- Paste your token
- Click Validate to verify
- Save the provider
API Usage
Model Names
{
"model": "Kimi-K2.5",
"messages": [...]
}Model Mapping Recommendations
| Model Name | Actual Model | Description |
|---|---|---|
Kimi-K2.5-Think | Kimi-K2.5 | Deep thinking model |
Kimi-K2.5-Search | Kimi-K2.5 | Web search model |
Kimi-K2.5-Think-Search | Kimi-K2.5 | Deep thinking + Web search |
Rate Limits
- Normal accounts: ~30 long-text conversations per 3 hours
- Short text conversations are unlimited
- Free accounts do not support deep thinking (official limitation)