open_responses library

OpenResponses API client for Dart.

Provides a unified, type-safe interface for interacting with multiple LLM providers through the OpenResponses specification.

Getting Started

import 'package:open_responses/open_responses.dart';

final client = OpenResponsesClient(
  config: OpenResponsesConfig(
    authProvider: BearerTokenProvider('your-api-key'),
  ),
);

final response = await client.responses.create(
  CreateResponseRequest(
    model: 'gpt-4o',
    input: ResponseTextInput('Hello, world!'),
  ),
);

print(response.outputText);

client.close();

Using with Different Providers

OpenAI (Default)

final client = OpenResponsesClient(
  config: OpenResponsesConfig(
    authProvider: BearerTokenProvider(Platform.environment['OPENAI_API_KEY']!),
  ),
);

Ollama (Local)

final client = OpenResponsesClient(
  config: OpenResponsesConfig(
    baseUrl: 'http://localhost:11434/v1',
    // No auth needed for local Ollama
  ),
);

Learn More

OpenResponses Specification

Classes

Annotation: Annotation for output content (citations, etc.).
AssistantMessageContent: Content for an assistant message.
AssistantMessageItem: An assistant message item.
AssistantMessagePartsContent: List of output content parts for an assistant message.
AssistantMessageTextContent: Simple text content for an assistant message.
AuthCredentials: Authentication credentials returned by AuthProvider.
AuthProvider: Authentication provider interface for the OpenResponses client.
BearerTokenCredentials: Bearer token authentication credentials.
BearerTokenProvider: Simple bearer token provider with static credentials.
ContentPartAddedEvent: Event indicating a content part was added.
ContentPartDoneEvent: Event indicating a content part is done.
CreateResponseRequest: Request to create a response.
DeveloperMessageItem: A developer message item.
ErrorEvent: Error event during streaming.
ErrorPayload: Error information from a failed response.
FileCitation: File citation annotation.
FilePathAnnotation: File path annotation.
FunctionCallArgumentsDeltaEvent: Event with function call arguments delta.
FunctionCallArgumentsDoneEvent: Event indicating function call arguments are done.
FunctionCallItem: A function call item.
FunctionCallOutput: The output of a function call.
FunctionCallOutputContent: A list of content items output from a function call.
FunctionCallOutputItem: A function call output item.
FunctionCallOutputItemResponse: A function call output item in the response.
FunctionCallOutputString: A string output from a function call.
FunctionTool: A function tool.
IncompleteDetails: Details about why a response was incomplete.
InputContent: Input content for messages.
InputFileContent: File content via URL, file ID, or base64-encoded data.
InputImageContent: Image content via URL.
InputTextContent: Text content.
InputTokensDetails: Detailed breakdown of input tokens.
InputVideoContent: Video content via URL.
Item: Input item for a response request.
ItemReference: Reference to a previously created item.
JsonObjectFormat: JSON object response format.
JsonSchemaFormat: JSON schema response format.
LogProb: Log probability information for a token.
McpTool: A Model Context Protocol (MCP) tool.
MessageItem: A message item in a conversation.
MessageOutputItem: A message output item.
NoAuthCredentials: No authentication credentials.
NoAuthProvider: Provider that returns no authentication credentials.
OpenResponsesClient: Dart client for the OpenResponses API.
OpenResponsesConfig: Configuration for the OpenResponses client.
OutputContent: Output content from model.
OutputItem: Output item from a response.
OutputItemAddedEvent: Event indicating an output item was added.
OutputItemDoneEvent: Event indicating an output item is done.
OutputTextAnnotationAddedEvent: Event indicating an annotation was added.
OutputTextContent: Text output content.
OutputTextDeltaEvent: Event with a text delta.
OutputTextDoneEvent: Event indicating output text is done.
OutputTokensDetails: Detailed breakdown of output tokens.
ReasoningConfig: Configuration for reasoning models.
ReasoningDeltaEvent: Event with reasoning delta.
ReasoningDoneEvent: Event indicating reasoning is done.
ReasoningInputItem: A reasoning item used as input for multi-turn conversations.
ReasoningItem: A reasoning item from reasoning models.
ReasoningSummaryContent: Content within a reasoning summary.
ReasoningSummaryDeltaEvent: Event with reasoning summary delta.
ReasoningSummaryDoneEvent: Event indicating reasoning summary is done.
ReasoningSummaryPartAddedEvent: Event indicating a reasoning summary part was added.
ReasoningSummaryPartDoneEvent: Event indicating a reasoning summary part is done.
ReasoningTextContent: Reasoning text content from reasoning models.
RefusalContent: Refusal content when model declines to respond.
RefusalDeltaEvent: Event with a refusal delta.
RefusalDoneEvent: Event indicating refusal is done.
RequestMetadata: Metadata about the HTTP request that caused an error.
ResponseCompletedEvent: Event indicating a response completed successfully.
ResponseCreatedEvent: Event indicating a response was created.
ResponseFailedEvent: Event indicating a response failed.
ResponseIncompleteEvent: Event indicating a response was incomplete.
ResponseInProgressEvent: Event indicating a response is in progress.
ResponseInput: Type-safe input for a response request.
ResponseItemsInput: A list of items input.
ResponseMetadata: Metadata about the HTTP response that caused an error.
ResponseQueuedEvent: Event indicating a response was queued.
ResponseResource: A response resource from the API.
ResponsesResource: Resource for the Responses API.
ResponseStream: A streaming response with event-based callbacks.
ResponseTextInput: A simple text input.
RetryPolicy: Retry policy configuration.
ServiceTier: Service tier for request processing.
SpecificFunctionChoice: A specific function that can be selected.
SpecificToolChoice: A specific tool that can be selected.
StreamingEvent: Server-sent event for response streaming.
StreamingEventAccumulator: Accumulates streaming events into progressive state snapshots.
StreamingEventAccumulatorSnapshot: Accumulates streaming events into a progressive state.
StreamOptions: Options that control streamed response behavior.
SummaryTextContent: Summary text content from reasoning models.
SystemMessageItem: A system message item.
TextConfig: Configuration for text output.
TextFormat: Text output format.
TextResponseFormat: Plain text response format.
Tool: Tool definition for function calling.
ToolChoice: Tool choice specification.
ToolChoiceAllowedTools: Model can only call the specified tools.
ToolChoiceAuto: Model automatically decides whether to call tools.
ToolChoiceFunction: Model must call the specified function.
ToolChoiceNone: No tool should be called.
ToolChoiceRequired: Model must call at least one tool.
TopLogProb: Top alternative token with log probability.
UnknownEvent: An unrecognized streaming event type.
UrlCitation: URL citation annotation.
Usage: Token usage statistics.
UserMessageContent: Content for a user message.
UserMessageItem: A user message item.
UserMessagePartsContent: List of input content parts for a user message.
UserMessageTextContent: Simple text content for a user message.

Enums

AbortionStage: Stage at which request was aborted.
FunctionCallStatus: The status of a function call output.
ImageDetail: Image detail level for multimodal inputs.
Include: Additional data to include in the response.
ItemStatus: The status of an item.
MessageRole: The role of a message in a conversation.
ReasoningEffort: Reasoning effort level for reasoning models.
ReasoningSummary: Reasoning summary verbosity level.
ResponseStatus: The status of a response.
ToolChoiceMode: Tool choice mode for allowed tools.
Truncation: Truncation strategy for long inputs.
Verbosity: Verbosity level for output.

Extensions

ResponseResourceExtensions on ResponseResource: Convenience extensions for ResponseResource.
StreamingEventExtensions on StreamingEvent: Convenience extensions for StreamingEvent.
StreamingEventsExtensions on Stream<StreamingEvent>: Convenience extensions for streaming responses.

Constants

unsetCopyWithValue → const Object: Sentinel value for copyWith to distinguish null from unset.

Functions

listHash<T>(List<T>? list) → int: Content-based hash code for a nullable list.
listOfMapsDeepEqual(List<Map<String, dynamic>>? a, List<Map<String, dynamic>>? b) → bool: Compares two lists of maps for deep equality (handles nested maps and lists).
listOfMapsEqual<K, V>(List<Map<K, V>>? a, List<Map<K, V>>? b) → bool: Compares two lists of maps for shallow equality.
listOfMapsHash<K, V>(List<Map<K, V>>? list) → int: Content-based hash code for a nullable list of maps.
listOfMapsHashCode(List<Map<String, dynamic>>? list) → int: Computes a deep hash code for a list of maps.
listsEqual<T>(List<T>? a, List<T>? b) → bool: Compares two lists for equality.
mapDeepHashCode(Map<String, dynamic>? map) → int: Computes a deep hash code for a map (handles nested maps and lists).
mapHash<K, V>(Map<K, V>? map) → int: Content-based hash code for a nullable map (order-independent).
mapsDeepEqual(Map<String, dynamic>? a, Map<String, dynamic>? b) → bool: Compares two maps for deep equality (handles nested maps and lists).
mapsEqual<K, V>(Map<K, V>? a, Map<K, V>? b) → bool: Compares two maps for equality.

Exceptions / Errors

AbortedException: Exception thrown when a request is aborted by user.
ApiException: Exception for HTTP/API errors.
AuthenticationException: Exception for authentication errors (HTTP 401).
OpenResponsesException: Base sealed class for all OpenResponses exceptions.
RateLimitException: Exception for rate limit (429) errors.
TimeoutException: Exception for request timeouts.
ValidationException: Exception for validation errors (HTTP 400).

open_responses library

Getting Started

Using with Different Providers

OpenAI (Default)

Ollama (Local)

Learn More

Classes

Enums

Extensions

Constants

Functions

Exceptions / Errors

open_responses package

open_responses library