langchain-ai/open_deep_research

Project overview

An open-source deep research agent that coordinates LLMs, search APIs, and MCP servers to produce automated research reports. It is configurable across many model providers and search tools and can be run locally or deployed to the Open Agent Platform.

Project type

AI Agent · RAG · AI Search

Use cases

Search & Research

Deployment

Refer to project documentation

License

MIT

Best for

• Developers and researchers needing an open source agent configurable across multiple model providers and search tools.
• Users looking to evaluate deep research agent performance against PhD-level benchmarks.

Key capabilities

• Automated deep research agent that produces research reports by coordinating LLMs, search APIs, and MCP servers
• Supports a wide range of LLM providers via init_chat_model() API for summarization, research, compression, and final report generation
• Supports a wide range of search tools including Tavily search API, native web search for Anthropic and OpenAI, and full MCP compatibility
• Has full MCP compatibility allowing use of Model Context Protocol servers
• Configured for evaluation with Deep Research Bench benchmark of 100 PhD-level research tasks using LLM-as-a-judge
• Launches locally with LangGraph server and provides a web UI for interacting with the agent
• Can be deployed to Open Agent Platform for non-technical users to build and configure their own agents

Limitations and risks

• Selected models must support structured outputs and tool calling.
• Running comprehensive evaluation across 100 examples can cost $20-$100 depending on model selection.
• The project depends on external services for core functionality, such as requiring the Tavily search API as the default search tool.

Getting started

• Setup requires cloning the repository, activating a Python 3.11 virtual environment, installing dependencies with uv sync, and setting up a .env file with API keys for models and search tools.

Evidence and sources

• README: This is a simple, configurable, fully open source deep research agent that works across many model providers, search tools, and MCP servers.
• README: git clone https://github.com/langchain-ai/open_deep_research.git
• README: Achieved #6 ranking on the Deep Research Bench Leaderboard with an overall score of 0.4344.
• README: Open Deep Research supports a wide range of LLM providers via the init_chat_model() API. It uses LLMs for a few different tasks.
• README: Open Deep Research supports a wide range of search tools. By default it uses the Tavily search API. Has full MCP compatibility and work native web search for Anthropic and OpenAI.

Reviewed at: 2026-07-21T22:29:16Z