Struct OpenAICodexResponsesProvider 

pub struct OpenAICodexResponsesProvider { /* private fields */ }

Available on crate feature openai-codex only.

OpenAI Codex / ChatGPT subscription provider.

This provider uses the ChatGPT Codex backend (/backend-api/codex/responses) and requires an OAuth access token obtained from the ChatGPT Plus/Pro login flow.

Implementations

impl OpenAICodexResponsesProvider

pub fn new(api_key: String, model: String) -> OpenAICodexResponsesProvider

Create a new OpenAI Codex provider.

pub fn with_base_url( api_key: String, model: String, base_url: String, ) -> OpenAICodexResponsesProvider

Create a provider with a custom base URL.

pub fn gpt53_codex(api_key: String) -> OpenAICodexResponsesProvider

Create a provider using GPT-5.3-Codex (latest codex model).

pub fn codex(api_key: String) -> OpenAICodexResponsesProvider

Create a provider using the latest Codex model.

pub fn gpt54(api_key: String) -> OpenAICodexResponsesProvider

Create a provider using GPT-5.4 (frontier reasoning with 1.05M context).

pub const fn with_thinking( self, thinking: ThinkingConfig, ) -> OpenAICodexResponsesProvider

Set the provider-owned thinking configuration for this model.

pub fn with_account_id( self, account_id: impl Into<String>, ) -> OpenAICodexResponsesProvider

Set a known ChatGPT account id, avoiding JWT decoding on each request.

pub fn with_reasoning_effort( self, effort: ReasoningEffort, ) -> OpenAICodexResponsesProvider

Set the reasoning effort level.

pub const fn with_websockets_disabled( self, disabled: bool, ) -> OpenAICodexResponsesProvider

Force the HTTP transport, skipping the WebSocket path entirely.

In a WebSocket-hostile environment (a corporate proxy / firewall that black-holes the wss upgrade) the WebSocket-first transport stalls for up to the connect + warmup timeout budget on every fresh session before falling back to HTTP. An operator who knows their network cannot do wss can set this to skip the penalty entirely. The OPENAI_CODEX_DISABLE_WEBSOCKETS environment variable does the same without a code change.

Trait Implementations

impl Clone for OpenAICodexResponsesProvider

fn clone(&self) -> OpenAICodexResponsesProvider

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl LlmProvider for OpenAICodexResponsesProvider

fn chat<'life0, 'async_trait>( &'life0 self, request: ChatRequest, ) -> Pin<Box<dyn Future<Output = Result<ChatOutcome, Error>> + Send + 'async_trait>>

where 'life0: 'async_trait, OpenAICodexResponsesProvider: 'async_trait,

Non-streaming chat completion.

fn chat_stream( &self, request: ChatRequest, ) -> Pin<Box<dyn Stream<Item = Result<StreamDelta, Error>> + Send + '_>>

Streaming chat completion.

fn model(&self) -> &str

fn provider(&self) -> &'static str

fn configured_thinking(&self) -> Option<&ThinkingConfig>

Provider-owned thinking configuration, if any.

fn list_models<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = Result<Vec<ModelInfo>, Error>> + Send + 'async_trait>>

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

List the models the provider currently exposes, queried live from the provider’s own model-listing endpoint.

fn capabilities(&self) -> Option<&'static ModelCapabilities>

Canonical capability metadata for this provider/model, if known.

fn validate_thinking_config( &self, thinking: Option<&ThinkingConfig>, ) -> Result<(), Error>

Validate a thinking configuration against the provider/model capabilities.

fn resolve_thinking_config( &self, request_thinking: Option<&ThinkingConfig>, ) -> Result<Option<ThinkingConfig>, Error>

Resolve the effective thinking configuration for a request.

fn default_max_tokens(&self) -> u32

Default maximum output tokens for this provider/model when the caller does not explicitly override AgentConfig.max_tokens.

fn structured_output_support(&self) -> StructuredOutputSupport

How this provider satisfies a structured-output (ResponseFormat) request.