Crate erio_llm_client 

erio-llm-client

erio-llm-client is Erio’s provider abstraction for chat/completion style LLM calls. It includes request/response models, error handling, and an OpenAI- compatible provider implementation.

Use this crate to keep provider integration behind a trait while sharing a single request/response shape across agents and workflows.

Quickstart

use erio_core::Message;
use erio_llm_client::{CompletionRequest, LlmProvider, OpenAiProvider};

async fn run() -> Result<(), Box<dyn std::error::Error>> {
    let provider = OpenAiProvider::new("https://api.openai.com/v1", "YOUR_API_KEY");

    let request = CompletionRequest::new("gpt-4o-mini")
        .message(Message::system("You are concise."))
        .message(Message::user("Say hello in one sentence."));

    let _response = provider.complete(request).await?;
    Ok(())
}

API tour

• Provider types: LlmProvider, OpenAiProvider
• Request types: CompletionRequest, ToolDefinition
• Response types: CompletionResponse, StreamChunk, Usage
• Error type: LlmError
• Modules: openai, provider, request, response, error

Related crates

• Uses erio-core::Message as the canonical message format.
• Integrates with erio-tools by passing ToolDefinition in requests.
• Docs: https://docs.rs/erio-llm-client
• Source: https://github.com/NomanworkGroup/erio/tree/main/crates/llm-client

Compatibility

• MSRV: Rust 1.93
• License: Apache-2.0 Erio LLM Client - LLM provider abstraction and adapters for the agent runtime.

Re-exports

pub use error::LlmError;

pub use openai::OpenAiProvider;

pub use provider::LlmProvider;

pub use request::CompletionRequest;

pub use request::ToolDefinition;

pub use response::CompletionResponse;

pub use response::StreamChunk;

pub use response::Usage;

Modules

error

LLM-specific error types.

openai

OpenAI-compatible LLM provider.

provider

LLM provider trait definition.

request

Completion request types for LLM providers.

response

Completion response types for LLM providers.