Implementation:Elevenlabs Elevenlabs python PromptAgentApiModelInput

Field	Value
source	Elevenlabs_Elevenlabs_python
domains	Conversational AI, Agent Prompt, LLM Configuration, RAG, Tools
last_updated	2026-02-15

Overview

Description

PromptAgentApiModelInput is a Pydantic model that defines the prompt and LLM configuration for a conversational AI agent in the ElevenLabs platform. It is one of the most feature-rich configuration models in the SDK, encapsulating the system prompt, LLM selection (including custom LLM support), reasoning parameters, tool integrations (built-in tools, custom tools, MCP servers), knowledge base connections, RAG configuration, and backup LLM cascading settings.

This model is auto-generated by Fern from the ElevenLabs API definition and inherits from UncheckedBaseModel. It is used as input when creating or updating agent prompt configurations.

Usage

PromptAgentApiModelInput is used when creating or updating the prompt configuration of a conversational AI agent. It controls how the agent's underlying LLM processes conversations, what tools are available, and how knowledge retrieval is configured.

Code Reference

Source Location

src/elevenlabs/types/prompt_agent_api_model_input.py

Class Signature

class PromptAgentApiModelInput(UncheckedBaseModel):
    ...

Import Statement

from elevenlabs.types import PromptAgentApiModelInput

I/O Contract

Field	Type	Required	Description
prompt	Optional[str]	No	The prompt for the agent.
llm	Optional[Llm]	No	The LLM to query with the prompt and the chat history. If using data residency, the LLM must be supported in the data residency environment.
reasoning_effort	Optional[LlmReasoningEffort]	No	Reasoning effort of the model. Only available for some models.
thinking_budget	Optional[int]	No	Max number of tokens used for thinking. Use 0 to turn off if supported by the model.
temperature	Optional[float]	No	The temperature for the LLM.
max_tokens	Optional[int]	No	If greater than 0, maximum number of tokens the LLM can predict.
tool_ids	Optional[List[str]]	No	A list of IDs of tools used by the agent.
built_in_tools	Optional[BuiltInToolsInput]	No	Built-in system tools to be used by the agent.
mcp_server_ids	Optional[List[str]]	No	A list of MCP server IDs to be used by the agent.
native_mcp_server_ids	Optional[List[str]]	No	A list of Native MCP server IDs to be used by the agent.
knowledge_base	Optional[List[KnowledgeBaseLocator]]	No	A list of knowledge bases to be used by the agent.
custom_llm	Optional[CustomLlm]	No	Definition for a custom LLM if LLM field is set to 'CUSTOM_LLM'.
ignore_default_personality	Optional[bool]	No	Whether to remove the default personality lines from the system prompt.
rag	Optional[RagConfig]	No	Configuration for RAG.
timezone	Optional[str]	No	Timezone for displaying current time in system prompt (e.g., 'America/New_York', 'Europe/London', 'UTC').
backup_llm_config	Optional[PromptAgentApiModelInputBackupLlmConfig]	No	Configuration for backup LLM cascading. Can be disabled, use system defaults, or specify custom order.
cascade_timeout_seconds	Optional[float]	No	Time in seconds before cascading to backup LLM. Must be between 2 and 15 seconds.
tools	Optional[List[PromptAgentApiModelInputToolsItem]]	No	A list of tools that the agent can use over the course of the conversation (deprecated, use tool_ids instead).

Usage Examples

Basic Prompt Configuration

from elevenlabs.types import PromptAgentApiModelInput

prompt_config = PromptAgentApiModelInput(
    prompt="You are a helpful customer support agent for Acme Corp.",
    temperature=0.7,
    max_tokens=1024,
)

Advanced Configuration with RAG and Tools

from elevenlabs.types import PromptAgentApiModelInput, RagConfig, BuiltInToolsInput, CustomLlm

prompt_config = PromptAgentApiModelInput(
    prompt="You are a knowledgeable assistant with access to company documentation.",
    temperature=0.5,
    tool_ids=["tool_abc123", "tool_def456"],
    built_in_tools=BuiltInToolsInput(
        end_call=SystemToolConfigInput(enabled=True),
    ),
    rag=RagConfig(
        enabled=True,
        max_documents_length=5000,
    ),
    timezone="America/New_York",
    cascade_timeout_seconds=5.0,
)

Using a Custom LLM

from elevenlabs.types import PromptAgentApiModelInput, CustomLlm

prompt_config = PromptAgentApiModelInput(
    prompt="You are a helpful assistant.",
    custom_llm=CustomLlm(
        url="https://my-llm-endpoint.example.com/v1/chat/completions",
        model_id="my-custom-model",
    ),
)

Related Pages

• Elevenlabs_Elevenlabs_python_CustomLlm - Custom LLM configuration referenced by this model
• Elevenlabs_Elevenlabs_python_RagConfig - RAG configuration referenced by this model
• Elevenlabs_Elevenlabs_python_BuiltInToolsInput - Built-in tools configuration referenced by this model
• Elevenlabs_Elevenlabs_python_AgentConfig - Core agent configuration that contains the prompt
• Elevenlabs_Elevenlabs_python_GetAgentResponseModel - Full agent response model