Implementation:CrewAIInc CrewAI ContextualAI Parse Tool

Knowledge Sources	CrewAI
Domains	Tools, Document_Parsing, ContextualAI
Last Updated	2026-02-11 00:00 GMT

Overview

Concrete tool for parsing documents using Contextual AI's advanced document parser provided by CrewAI.

Description

The ContextualAIParseTool class extends BaseTool to parse complex documents including tables, figures, and multi-column layouts via Contextual AI's REST API (https://api.contextual.ai/v1/parse). The _run method executes a three-phase process: first, it submits a parse job with the document file and configuration parameters; second, it polls the job status every 5 seconds until completion or failure; third, it retrieves the parsed results with specified output types. Configuration options include parse_mode (standard), figure_caption_mode (concise), enable_document_hierarchy, page_range, and output_types (defaulting to markdown-per-page). The tool uses direct REST requests rather than an SDK, with bearer token authentication.

Usage

Use this tool when CrewAI agents need to preprocess complex PDF documents before ingesting them into RAG systems, enabling semantic search over structured content extracted from challenging document formats.

Code Reference

Source Location

• Repository: CrewAI
• File: lib/crewai-tools/src/crewai_tools/tools/contextualai_parse_tool/contextual_parse_tool.py
• Lines: 1-108

Signature

class ContextualAIParseTool(BaseTool):
    name: str = "Contextual AI Document Parser"
    description: str = "Parse documents using Contextual AI's advanced document parser"
    args_schema: type[BaseModel] = ContextualAIParseSchema
    api_key: str

    def _run(self, file_path: str, parse_mode: str = "standard",
             figure_caption_mode: str = "concise",
             enable_document_hierarchy: bool = True,
             page_range: str | None = None,
             output_types: list[str] | None = None) -> str: ...

Import

from crewai_tools import ContextualAIParseTool

I/O Contract

Inputs

Name	Type	Required	Description
api_key	str	Yes	Contextual AI API key (constructor)
file_path	str	Yes	Path to the document to parse
parse_mode	str	No	Parsing mode (default "standard")
figure_caption_mode	str	No	Figure caption mode (default "concise")
enable_document_hierarchy	bool	No	Enable document hierarchy extraction (default True)
page_range	str or None	No	Page range to parse (e.g., "0-5")
output_types	list[str] or None	No	List of output types (default ["markdown-per-page"])

Outputs

Name	Type	Description
_run() returns	str	JSON string containing parsed document content with the specified output types

Usage Examples

Basic Usage

from crewai_tools import ContextualAIParseTool

tool = ContextualAIParseTool(api_key="your-api-key")
result = tool.run(
    file_path="./documents/report.pdf",
    parse_mode="standard",
    page_range="0-10",
    output_types=["markdown-per-page"]
)

Related Pages

• Principle:CrewAIInc_CrewAI_Built_In_Tool_Selection