Scenario ยท verified 2026-09-15

Best agent frameworks for research agents

Research agents need retrieval context, source tracking, browser automation, and multi-step synthesis boundaries.

agent-framework

LlamaIndex

Data-oriented agent and workflow framework for building LLM agents over private data, tools, retrieval, and workflows.

tool-callingretrieval-contextworkflow-orchestrationmulti-agent-workflow

agent-framework

LangChain / LangGraph

Agent and graph runtime ecosystem for building tool-using, stateful, and observable LLM applications.

tool-callinggraph-runtimetracingevaluation

agent-framework

AutoGen

Microsoft open-source framework for multi-agent conversation patterns, agent teams, and collaborative task solving.

multi-agent-workflowworkflow-orchestrationtool-callingevaluation

mcp-server

Hugging Face MCP Server

MCP server entry for Hugging Face Hub model, dataset, and agent integration workflows.

model-hub-contextdataset-contextagent-integration

mcp-server

Playwright MCP Server

MCP server from Microsoft for browser automation and testing workflows backed by Playwright.

browser-automationtesting-contexttool-use

Ranking signals

SignalWeightRationale
Source provenance5Research workflows need inspectable sources and citation boundaries.
Retrieval and browsing5Research agents often combine retrieval context with browser automation.
Multi-step synthesis4Research outputs need orchestration across collection, filtering, and synthesis steps.

Source boundary

Scenario recommendation is derived from retrieval, browser automation, multi-agent workflow, and source verification metadata in this graph.

How should teams choose between AutoGen and CrewAI?

Choose AutoGen for conversational multi-agent experiments and Microsoft ecosystem alignment. Choose CrewAI for role-based crews, task delegation, and team-style workflow automation.

What should a research-agent stack verify before implementation?

Verify retrieval sources, browser automation boundaries, citation handling, and whether the selected framework can keep source provenance visible across multi-step synthesis.