Struct OpenAI 

pub struct OpenAI { /* private fields */ }

OpenAI model backed by the Responses API by default, with legacy chat.completions support for compatibility.

Implementations

impl OpenAI

pub fn new(api_key: impl Into<String>) -> OpenAI

Create a new client using the provided API key and default model.

pub fn deepseek(api_key: impl Into<String>) -> OpenAI

Create a client configured for Deepseek’s OpenAI-compatible endpoint.

pub fn openrouter(api_key: impl Into<String>) -> OpenAI

Create a client configured for OpenRouter’s OpenAI-compatible endpoint.

pub fn builder(api_key: impl Into<String>) -> Builder

Start building an OpenAI client with custom configuration.

pub fn with_model(self, model: impl Into<String>) -> OpenAI

Override the default chat model in-place.

pub fn with_base_url(self, base_url: impl Into<String>) -> OpenAI

Override the REST base URL (useful for OpenAI-compatible endpoints).

pub fn with_api(self, api: ApiKind) -> OpenAI

Select which OpenAI API to call.

pub fn with_responses_api(self) -> OpenAI

Use the recommended Responses API.

pub fn with_chat_completions_api(self) -> OpenAI

Use the legacy Chat Completions API (deprecated by OpenAI).

pub fn with_legacy_max_tokens(self, enabled: bool) -> OpenAI

Send deprecated max_tokens alongside max_completion_tokens for compatibility.

OpenAI deprecates max_tokens and it is incompatible with reasoning models.

pub fn with_embedding_model(self, model: impl Into<String>) -> OpenAI

Override the embeddings model identifier.

pub fn with_embedding_dimensions(self, dimensions: usize) -> OpenAI

Override the embedding dimension (defaults depend on model).

pub fn with_image_model(self, model: impl Into<String>) -> OpenAI

Override the image generation model identifier.

pub fn with_audio_model(self, model: impl Into<String>) -> OpenAI

Override the default text-to-speech model identifier.

pub fn with_audio_voice(self, voice: impl Into<String>) -> OpenAI

Set the default text-to-speech voice (e.g., alloy).

pub fn with_audio_format(self, format: impl Into<String>) -> OpenAI

Set the preferred audio format (mp3, wav, flac).

pub fn with_transcription_model(self, model: impl Into<String>) -> OpenAI

Override the transcription model identifier.

pub fn with_moderation_model(self, model: impl Into<String>) -> OpenAI

Override the moderation model identifier.

Trait Implementations

impl AudioGenerator for OpenAI

type Error = OpenAIError

The error type returned by this generator.

fn generate( &self, prompt: &str, ) -> impl Stream<Item = Result<Vec<u8>, <OpenAI as AudioGenerator>::Error>> + Send

Generates audio from text prompt.

impl AudioTranscriber for OpenAI

type Error = OpenAIError

The error type returned by this transcriber.

fn transcribe( &self, audio: &[u8], ) -> impl Stream<Item = Result<String, <OpenAI as AudioTranscriber>::Error>> + Send

Transcribes audio data to text.

impl Clone for OpenAI

fn clone(&self) -> OpenAI

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for OpenAI

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

Formats the value using the given formatter.

impl EmbeddingModel for OpenAI

fn dim(&self) -> usize

Returns the embedding vector dimension.

fn embed( &self, text: &str, ) -> impl Future<Output = Result<Vec<f32>, Error>> + Send

Converts text to an embedding vector.

impl From<OpenAI> for CloudEmbedder

fn from(client: OpenAI) -> Self

Converts to this type from the input type.

impl From<OpenAI> for CloudProvider

fn from(client: OpenAI) -> Self

Converts to this type from the input type.

impl ImageGenerator for OpenAI

type Error = OpenAIError

The error type returned by the image generator.

fn create( &self, prompt: Prompt, size: Size, ) -> impl Stream<Item = Result<Vec<u8>, <OpenAI as ImageGenerator>::Error>> + Send

Create an image from a prompt and a specified size.

fn edit( &self, prompt: Prompt, mask: &[u8], ) -> impl Stream<Item = Result<Vec<u8>, <OpenAI as ImageGenerator>::Error>> + Send

Edit an image using a prompt and a mask.

impl LanguageModel for OpenAI

type Error = OpenAIError

The error type returned by this language model.

fn respond( &self, request: LLMRequest, ) -> impl Stream<Item = Result<Event, <OpenAI as LanguageModel>::Error>> + Send

Generates a streaming response to a conversation.

fn complete( &self, prefix: &str, ) -> impl Stream<Item = Result<Event, <OpenAI as LanguageModel>::Error>> + Send

Completes given text prefix.

fn profile(&self) -> impl Future<Output = Profile> + Send

Returns model profile and capabilities.

fn respond_with_tools( &self, request: LLMRequestWithTools<'_>, ) -> impl Stream<Item = Result<Event, Self::Error>> + Send

Generates a streaming response with a mutable tool registry.

fn generate<T>( &self, request: LLMRequest, ) -> impl Future<Output = Result<T, GenerateError<Self::Error>>> + Send

where T: JsonSchema + DeserializeOwned + 'static,

Generates structured output conforming to JSON schema.

fn summarize( &self, text: &str, ) -> impl Stream<Item = Result<Event, Self::Error>> + Send

Summarizes text.

fn categorize<T>( &self, text: &str, ) -> impl Future<Output = Result<T, GenerateError<Self::Error>>> + Send

where T: JsonSchema + DeserializeOwned + 'static,

Categorizes text.

impl Moderation for OpenAI

type Error = OpenAIError

The error type returned by moderation operations.

fn moderate( &self, content: &str, ) -> impl Future<Output = Result<ModerationResult, <OpenAI as Moderation>::Error>> + Send

Moderates the provided content and returns a result asynchronously.