dartantic_ai
library
Classes
-
Agent
-
An agent that can run prompts through an AI model and return responses.
-
AgentResponseFor<T>
-
A response from an agent.
-
DataPart
-
A data part of a message's content, such as an image or video.
-
GeminiModel
-
Implementation of Model that uses Google's Gemini API.
-
GeminiProvider
-
Provider for Google's Gemini AI models.
-
LinkPart
-
A link part of a message's content, representing a URL.
-
McpClient
-
Configuration for connecting to an MCP (Model Context Protocol) server.
-
Message
-
Represents a message in a conversation, consisting of a role and a list of
content parts.
-
Model
-
Abstract interface for AI model implementations.
-
ModelInfo
-
Information about an available AI model.
-
ModelSettings
-
Settings used to configure a model.
-
OpenAiModel
-
Implementation of Model that uses OpenAI's API.
-
OpenAiProvider
-
Provider for OpenAI models.
-
Part
-
An abstract base class for a part of a message's content.
-
Provider
-
Abstract interface for AI model providers.
-
ProviderSettings
-
Settings used to configure a provider.
-
ProviderTable
-
Manages the mapping between provider family names and provider factories.
-
TextPart
-
A text part of a message's content.
-
Tool
-
Represents a tool that can be called by an agent to perform a specific
function.
-
ToolPart
-
A tool call part of a message's content, representing a call to an external
tool or function.
Enums
-
EmbeddingType
-
The type of embedding to generate.
-
McpServerKind
-
Represents the type of MCP server connection.
-
MessageRole
-
Enum representing the role of a message in a conversation.
-
ModelKind
-
The type of AI model based on its primary capability.
-
ProviderCaps
-
Defines capabilities providers can have.
-
ToolCallingMode
-
The mode in which the model will call tools.
-
ToolPartKind
-
Enum to distinguish between tool call and tool result parts.
