Crate miyabi_llm_openai 

OpenAI GPT SDK for Miyabi LLM

This crate provides OpenAI GPT API integration for the Miyabi LLM framework. It implements the LlmClient and LlmStreamingClient traits from miyabi-llm-core.

Features

• GPT-4o, GPT-4o-mini, GPT-4 Turbo support
• Tool/function calling
• Streaming responses via SSE
• Environment variable configuration

Example

use miyabi_llm_openai::OpenAIClient;
use miyabi_llm_core::{LlmClient, Message};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let client = OpenAIClient::from_env()?;
    let messages = vec![Message::user("Hello!")];
    let response = client.chat(messages).await?;
    println!("Response: {}", response);
    Ok(())
}

Structs

Message

Message in a conversation

OpenAIChoice

Choice in OpenAI response

OpenAIClient

OpenAI GPT client

OpenAIFunction

OpenAI function definition

OpenAIFunctionCall

Function call in OpenAI tool call

OpenAIMessage

OpenAI message format

OpenAIResponse

OpenAI API response

OpenAIResponseMessage

Response message in OpenAI choice

OpenAITool

OpenAI tool definition

OpenAIToolCall

Tool call in OpenAI response

OpenAIUsage

Token usage information

ToolCall

Tool call from LLM

ToolDefinition

Tool definition for LLM Function Calling

Enums

LlmError

Core LLM error type

Role

Message role

StreamEvent

Stream event type for more detailed streaming information

ToolCallResponse

Response from LLM with tool calling support

Traits

LlmClient

LLM Client trait - unified interface for all providers

LlmStreamingClient

LLM Streaming Client trait

Type Aliases

Result

Result type for LLM operations

StreamResponse

Type alias for streaming response