ai.z.openapi.api.chat

Interface ChatApi

public interface ChatApi

Chat Completions API for advanced GLM-4 series models Provides synchronous, asynchronous, and streaming chat completion capabilities Supports complex reasoning, long context processing (up to 128K tokens), and ultra-fast inference Features GLM-4-Plus, GLM-4-Air, GLM-4-Flash, and GLM-4-AllTools with specialized capabilities Optimized for Chinese and multilingual conversations with superior performance

Method Summary

Modifier and Type	Method and Description
io.reactivex.Single<ModelData>	createChatCompletion(ChatCompletionCreateParams request) Create a synchronous chat completion with immediate response Waits for the GLM-4 model to complete execution and returns the final result Supports complex reasoning, tool calling, function execution, and multi-modal understanding Features advanced capabilities like web search integration, code interpretation, and image analysis
io.reactivex.Single<ModelData>	createChatCompletionAsync(ChatCompletionCreateParams request) Create an asynchronous chat completion for long-running tasks Submits the request and returns immediately with a task ID for later result retrieval Ideal for complex reasoning tasks, long document processing, or batch operations Supports advanced GLM-4 models with extended context and computational requirements
retrofit2.Call<okhttp3.ResponseBody>	createChatCompletionStream(ChatCompletionCreateParams request) Create a streaming chat completion with real-time response Returns response content incrementally through Server-Sent Events (SSE) for immediate user feedback Optimized for interactive applications requiring low latency and progressive content delivery Supports all GLM-4 models with configurable streaming parameters and token-by-token generation
io.reactivex.Single<ModelData>	queryAsyncResult(String id) Query the result of an asynchronous chat completion task Retrieves the completion result or current status using the task ID from async request Provides detailed progress information for long-running tasks and complex reasoning operations Supports polling-based result retrieval with comprehensive status reporting

Method Detail

createChatCompletionStream

@Streaming
 @POST(value="chat/completions")
retrofit2.Call<okhttp3.ResponseBody> createChatCompletionStream(@Body
                                                                                                            ChatCompletionCreateParams request)

Create a streaming chat completion with real-time response Returns response content incrementally through Server-Sent Events (SSE) for immediate user feedback Optimized for interactive applications requiring low latency and progressive content delivery Supports all GLM-4 models with configurable streaming parameters and token-by-token generation

Parameters:

request - Chat completion parameters including model selection (glm-4-plus, glm-4-air, glm-4-flash), messages, temperature, top_p, max_tokens, and streaming settings

Returns:

Streaming response body with incremental content, usage statistics, and completion indicators

createChatCompletionAsync

@POST(value="async/chat/completions")
io.reactivex.Single<ModelData> createChatCompletionAsync(@Body
                                                                                               ChatCompletionCreateParams request)

Create an asynchronous chat completion for long-running tasks Submits the request and returns immediately with a task ID for later result retrieval Ideal for complex reasoning tasks, long document processing, or batch operations Supports advanced GLM-4 models with extended context and computational requirements

Parameters:

request - Chat completion parameters including model selection, messages, advanced reasoning settings, tools configuration, and processing options

Returns:

Task information with unique ID, estimated completion time, and processing status for asynchronous tracking

createChatCompletion

@POST(value="chat/completions")
io.reactivex.Single<ModelData> createChatCompletion(@Body
                                                                                    ChatCompletionCreateParams request)

Create a synchronous chat completion with immediate response Waits for the GLM-4 model to complete execution and returns the final result Supports complex reasoning, tool calling, function execution, and multi-modal understanding Features advanced capabilities like web search integration, code interpretation, and image analysis

Parameters:

request - Chat completion parameters including model selection, conversation messages, generation settings (temperature, top_p, max_tokens), tools configuration, and response format

Returns:

Complete chat completion response with generated content, usage statistics, tool call results, and reasoning traces

queryAsyncResult

@GET(value="async-result/{id}")
io.reactivex.Single<ModelData> queryAsyncResult(@Path(value="id")
                                                                                String id)

Query the result of an asynchronous chat completion task Retrieves the completion result or current status using the task ID from async request Provides detailed progress information for long-running tasks and complex reasoning operations Supports polling-based result retrieval with comprehensive status reporting

Parameters:

id - Unique task ID returned from asynchronous chat completion request

Returns:

Chat completion result with generated content, processing status, completion percentage, and any intermediate results or error information