gpt-researcher

To automate web research, initialize an autonomous agent with a query, execute the research task, and call the report generation function to produce a structured Markdown document with citations and references.

Autonomous research orchestration coordinates planning, execution, and publishing phases using a multi-agent workflow, parallelizing tasks across web and local data sources to gather information and compile comprehensive reports.

You can use MCP data sources and local retrievers for deep research by configuring them within the autonomous agent, allowing the system to gather and synthesize information from multiple parallelized data streams.

A configured LLM provider is required to conduct autonomous research, driving the planner-executor-publisher pattern to process gathered data and generate the final cited report.

The best way to monitor report generation progress is by configuring optional WebSocket streaming, which provides real-time updates as the research agent executes its planning and gathering tasks.

Autonomous report generation supports image generation alongside text, allowing the multi-agent workflow to include visual assets within the final structured Markdown output and citations.