Quick Start
1. Account Registration
The platform now supports online registration and login, and also supports obtaining Token quota through redemption codes. Please contact the platform administrator for how to get the redemption code. For detailed account registration, activation process and usage instructions, please refer to the Basic Usage documentation.
2. API Configuration
1. Get Your Token
- Log in to the dashboard
- Go to "API Keys" page
- Click "Add Key" to get your API Key
2. API Endpoint
The platform supports multiple API formats. Choose the corresponding address based on the model you use.
OpenAI Compatible Protocol
| Address Type | Full Address | Applicable Scenario |
|---|---|---|
| Base URL | https://ai-tokenhub.com/api/v1 | The base address to fill in when developing with an OpenAI-compatible SDK. The SDK will automatically append the endpoint path. |
| Chat Completions | https://ai-tokenhub.com/api/v1/chat/completions | Used when sending HTTP requests directly to call the chat dialogue interface. |
| Responses | https://ai-tokenhub.com/api/v1/responses | Used when sending HTTP requests directly to call the asynchronous task result query interface. |
Anthropic Protocol
| Address Type | Full Address | Applicable Scenario |
|---|---|---|
| Base URL | https://ai-tokenhub.com/api/anthropic | The base address to fill in when developing with the Anthropic native SDK. The SDK will automatically append the endpoint path. |
| Messages | https://ai-tokenhub.com/api/anthropic/v1/messages | Used when sending HTTP requests directly to call the chat dialogue interface. |
Multimodal Model APIs
Multimodal models use the Ark protocol and are not interchangeable with OpenAI/Anthropic chat interfaces (
/api/v1,/api/anthropic). The base address is/api/v2.
| Model Type | Full Address | Applicable Scenario |
|---|---|---|
| Image Models (Seedream Series) | https://ai-tokenhub.com/api/v2/images/generations | Send HTTP requests directly to call Seedream series image generation API |
| Audio Models (TTS) | https://ai-tokenhub.com/api/v2/tts/unidirectional | Send HTTP requests directly to call TTS model for unidirectional streaming speech synthesis |
| Video Models (Seedance Series) V3 | https://ai-tokenhub.com/api/v3/contents/generations/tasks | Send HTTP requests directly to create Seedance series video generation tasks (V3 API, asynchronous tasks, requires polling with task query API for results) |
| Video Models (Seedance Series) | https://ai-tokenhub.com/api/v2/contents/generations/tasks | Send HTTP requests directly to create Seedance series video generation tasks (asynchronous tasks, requires polling with task query API for results) |
3. Model Selection
Model names are listed in the "Models" section of the website's top navigation. Make sure to use the exact model name.
3. Testing & Verification
Use one of the following methods to verify your API configuration:
- Test online on the chat page (recommended if you want to integrate into your own chat software)
- Use the online debugging feature in the documentation (recommended for API-based application development)
- Call it in AI programming software (like Claude Code, Cline, etc.)
- Use API tools (like Postman) for interface testing
Request Example
For complete multi-language API request examples and debugging methods, please refer to the API Test documentation.