Struct OpenAiWorker 

pub struct OpenAiWorker { /* private fields */ }

OpenAI API worker (GPT-4, GPT-3.5-turbo-instruct, etc.)

Requires OPENAI_API_KEY environment variable.

Example

let worker = Arc::new(
    OpenAiWorker::new("gpt-3.5-turbo-instruct")?
        .with_max_tokens(512)
        .with_temperature(0.7)
);

Resilience

OpenAiWorker does not retry internally. Retry logic is handled by the pipeline’s inference stage. See ModelWorker for details.

Implementations

impl OpenAiWorker

pub fn new(model: impl Into<String>) -> Result<Self, OrchestratorError>

Create a new OpenAiWorker for the given model.

Reads the API key from the OPENAI_API_KEY environment variable and constructs an HTTP client configured for the OpenAI chat/completions endpoint. Default settings: 256 max tokens, temperature 0.7, 30 s timeout. All defaults are overridable via the builder methods.

Environment Variables

Variable	Required	Description
OPENAI_API_KEY	Yes	Bearer token sent with every request

Errors

Returns OrchestratorError::ConfigError if OPENAI_API_KEY is not set in the environment.

Examples

use tokio_prompt_orchestrator::OpenAiWorker;
use std::sync::Arc;

// OPENAI_API_KEY must be set in the environment.
let worker = Arc::new(
    OpenAiWorker::new("gpt-4o")
        .expect("OPENAI_API_KEY must be set")
        .with_max_tokens(512)
        .with_temperature(0.7),
);

pub fn with_max_tokens(self, max_tokens: u32) -> Self

Set maximum tokens to generate

pub fn with_temperature(self, temperature: f32) -> Self

Set temperature (0.0 – 2.0 for OpenAI).

Values outside [0.0, 2.0] are clamped and a WARN-level log line is emitted. The OpenAI API rejects values outside this range with HTTP 400.

pub fn with_timeout(self, timeout: Duration) -> Self

Set request timeout

pub fn with_base_url(self, url: impl Into<String>) -> Self

Override the API base URL.

Useful for OpenAI-compatible endpoints (Azure OpenAI, Groq, local proxies) and for pointing at a mock server in tests. Default: "https://api.openai.com/v1".

Trait Implementations

impl Debug for OpenAiWorker

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl ModelWorker for OpenAiWorker

fn infer<'life0, 'life1, 'async_trait>( &'life0 self, prompt: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<Vec<String>, OrchestratorError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Perform inference on the given prompt.

fn infer_stream<'life0, 'life1, 'async_trait>( &'life0 self, prompt: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<TokenStream, OrchestratorError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Stream inference tokens as they arrive from the provider.