Skip to main content

​
DeepResearchAgent

Defined in the deep_research_agent module.
Agent for performing deep research using multiple provider APIs. Supports:
  • OpenAI Deep Research: o3-deep-research, o4-mini-deep-research
  • Gemini Deep Research: deep-research-pro-preview
  • LiteLLM: Unified interface for OpenAI models
The provider is auto-detected based on the model name, or can be explicitly set.

​
Constructor

​
name
Optional
No description available.
​
instructions
Optional
No description available.
​
model
str
default:"'o3-deep-research'"
No description available.
​
provider
Optional
default:"'auto'"
No description available.
​
api_key
Optional
No description available.
​
base_url
Optional
No description available.
​
verbose
bool
default:"True"
No description available.
​
summary_mode
Literal
default:"'auto'"
No description available.
​
enable_web_search
bool
default:"True"
No description available.
​
enable_code_interpreter
bool
default:"False"
No description available.
​
enable_file_search
bool
default:"False"
No description available.
​
mcp_servers
Optional
No description available.
​
file_search_stores
Optional
No description available.
​
use_litellm
bool
default:"False"
No description available.
​
poll_interval
int
default:"10"
No description available.
​
max_wait_time
int
default:"3600"
No description available.

​
Methods

openai_client()

Get the synchronous OpenAI client (lazy initialization).

async_openai_client()

Get the asynchronous OpenAI client (lazy initialization).

gemini_client()

Get the Gemini client (lazy initialization).

research()

Perform a deep research query.

aresearch()

Async version of research().

follow_up()

Ask a follow-up question based on a previous research interaction.

clarify()

Generate clarifying questions for a research query.

rewrite_query()

Rewrite a research query to be more specific and detailed.

​
Usage

# OpenAI (default)
    agent = DeepResearchAgent(
        model="o3-deep-research",
        instructions="You are a professional researcher."
    )
    
    # Gemini
    agent = DeepResearchAgent(
        model="deep-research-pro",
        instructions="You are a professional researcher."
    )
    
    # Using LiteLLM
    agent = DeepResearchAgent(
        model="o3-deep-research",
        use_litellm=True
    )
    
    result = agent.research("Research the economic impact of AI on healthcare.")
    print(result.report)