Loading...
Loading...
Fully compatible with OpenAI API, supporting multiple mainstream LLMs
Common questions about model selection
For strong reasoning and coding, choose DeepSeek V4; for Chinese understanding, choose Zhipu GLM-5; for cost-effectiveness, choose Qwen Plus. You can also use the auto model for automatic optimal channel selection.
Pricing is per 1M tokens, split into input and output prices. Input tokens are what you send to the model, output tokens are what the model generates. Prices vary by model, see the pricing table above.
The auto model automatically load-balances across all available channels, selecting the most stable and fastest channel to process requests without manual model specification.
Yes, all models are relayed through Ciyuano and fully compatible with OpenAI Chat Completions API format. Just change base_url and api_key to use any OpenAI SDK.
If a model is temporarily unavailable, the system automatically retries other channels. You can also switch to the auto model for automatic channel selection.
Register for free credits, complete integration in 5 minutes. Supports DeepSeek, GLM, Qwen and more.
Free Registration