GPT Researcher
Open-source autonomous research agent with planner/execution multi-agent architecture and cited reports.
GPT Researcher is an open source autonomous research agent, released under the Apache 2.0 license and maintained by Assaf Elovic and contributors, that plans research tasks, runs parallel multi-source web search, validates sources, accumulates a research context and writes long-form reports with inline citations. A planner breaks a query into sub-questions, sub-agents crawl, scrape and summarize sources in parallel, and a publisher assembles the report. A quick search mode returns grounded snippets when a calling agent needs less.
The agent works with a wide range of model providers, from OpenAI, Anthropic, Google and Mistral to local models through Ollama, and with retrievers including Tavily, Bing, Google, DuckDuckGo, Exa, ArXiv and PubMed, custom MCP retrievers and local document stores. Reports come out as Markdown, PDF, DOCX or JSON, with live progress streamed over WebSocket or server-sent events.
It is built to be called by other agents and applications: it ships as a Python package, publishes an OpenAPI specification and an A2A agent card, and runs an official MCP server with deep research, quick search, report writing and source tools. Customers run it themselves, in Docker or as a service, with their own model and search provider keys.
Vendor details
Canonical URL
https://gptr.dev
Category
Agent infrastructure
Company status
independent
Use cases & customers
Deployment options
In practice
You need a sourced report on a topic and a single chatbot answer won't cut it. GPT Researcher plans sub-questions, crawls more than 20 sources in parallel, and aggregates a structured report with citations.
You worry the model is confidently making things up. GPT Researcher runs agents in parallel and cross-references many sources, which is how the project tackles hallucination and bias instead of trusting one call.
You want deeper research from inside the tools you already use. GPT Researcher ships an MCP server that lets an application like Claude run its full research loop and return a citation-backed report.
Sources & related URLs
Related / legacy domains
Agentic Index coverage score
7.0 / 14 capabilities · 50%
| Integrations & Tool Calling | Partial |
|---|---|
|
Crawler agents gather information from many search retrievers, including Tavily, Google, Bing, DuckDuckGo and Exa, and from local documents such as portable document format, Word and spreadsheet files. The agent also exposes itself over the Model Context Protocol. These tools serve research, and the agent takes no broad action in outside business systems. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
| Workflow Orchestration | Full |
|
A planner agent generates research questions, parallel execution agents scrape and summarize sources, and a publisher aggregates a cited report. The agents work autonomously. A multi agent variant is built on LangGraph, and a recursive deep research mode explores subtopics to a configurable depth and breadth. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
| Knowledge Grounding & RAG | Partial |
|
Beyond web search, GPT Researcher can research over local document stores. It keeps a vector store of the accumulated research context, with optional persistence across sessions, and draws customer documents in for each research task. GPT Researcher names no index it maintains over the customer's knowledge base outside a research run. SourceGPT Researcher, gptr.dev/llms-full.txtread 2026-09-21 |
|
| Human Oversight & Guardrails | Not documented |
|
The agent runs autonomously from a query to a report. GPT Researcher names no approval step, review gate or guardrail engine for a person to hold or approve its actions. SourceGPT Researcher, gptr.dev/llms-full.txtread 2026-09-21 |
|
| Security, Identity & Governance | Not documented |
|
The customer runs GPT Researcher as open source software, and the product has no authentication of its own. The only keys are the model and search provider keys the customer supplies. GPT Researcher names no access model, identity integration or attestation. SourceGPT Researcher, gptr.dev/llms-full.txtread 2026-09-21 |
|
| Observability & Auditability | Partial |
|
The agent tracks and cites the source of every summarized fact, so a report can be audited back to its origins. GPT Researcher names no full execution tracing, metrics or logging dashboard for agent runs. Sourcegptr.devread 2026-09-21 |
|
| Memory & State Persistence | Not documented |
|
Each research task runs and returns a report. GPT Researcher names no persistent agent memory or checkpoint layer that carries state across sessions. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
| Deployment & Data Residency | Full |
|
The project is open source under the Apache license. It deploys through Docker, a Python package or a FastAPI service, and it runs with local models through Ollama for fully private, offline operation. Customers host it themselves, so data stays wherever they run it. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
| Prebuilt Agents, Templates & Packs | Partial |
|
The project ships a ready to use research agent and a full suite of customization options for building tailor made, domain specific research agents. GPT Researcher lists no browsable marketplace of agents or templates. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
| Triggers & Channel Coverage | Partial |
|
Developers invoke the agent through a Python package, a representational state transfer application programming interface, a web interface, a command line, or the Model Context Protocol from assistants like Claude. GPT Researcher names no scheduled or event driven triggers. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
| Model Flexibility & Routing | Full |
|
The agent works with any large language model provider, including OpenAI, Anthropic, Groq, Google, Mistral and local models through Ollama. It assigns cheaper and stronger models to different roles, such as planning and execution, to optimize cost. Sourcepypi.org/project/gpt-researcherread 2026-09-21 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
GPT Researcher ships as a Python package, publishes an OpenAPI specification and an A2A agent card, and runs an official MCP server exposing deep research, quick search, report writing and source retrieval tools over stdio, SSE and streamable HTTP, so outside callers drive it directly. SourceGPT Researcher, gptr.dev/llms-full.txt and gptr.dev homepageread 2026-09-21 |
|
| Testing, Debugging & Optimization | Not documented |
|
Aggregating many sources improves research quality and reduces misinformation. GPT Researcher names no testing, debugging or evaluation engine for agents. Sourcegptr.devread 2026-09-21 |
|
| Browser & Computer Use | Partial |
|
Crawler agents browse and scrape web pages to gather information for each research question. The agent does not offer general autonomous browser or computer control as a product. Sourcegithub.com/assafelovic/gpt-researcherread 2026-09-21 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
GPT Researcher introduced Deep Research, an advanced recursive workflow for exploring topics with agentic depth. The update also moves the platform's Model Context Protocol (MCP) server to a dedicated repository (gptr-mcp). Additionally, it adds multi-agent assistants built with LangGraph and AG2 and an enhanced frontend for real-time progress tracking.
Bears on: MCP / tool calling / API
View sourcePricing
Free and open source
What is public
GPT Researcher is free, open source software under the Apache 2.0 license that the customer runs; the only costs are the model and search provider keys supplied at runtime.
Cost watchouts
Model provider and search provider usage are the customer's own costs and scale with the number and depth of research runs.
Related vendors
- AgentOps — Agent observability and debugging platform: open source SDKs trace…
- AgentPhone — AgentPhone gives the AI agents a team builds their own US or…
- Agno — Python agent framework and AgentOS runtime (formerly Phidata) for…
- AIsa — Resource and payment gateway for AI agents, with one key to 110+…
- Albato — Albato is a no code integration platform that sells an embedded,…
- AlphaBitCore — AI control plane for regulated financial firms: one gateway enforces…
Alternatives to GPT Researcher
The closest documented capability profiles to GPT Researcher among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Beam7.5 / 14Adds documented Security, Identity & Governance and Memory & State Persistence
- Jina AI5.0 / 14Adds documented Security, Identity & Governance
- Lemony7.0 / 14Adds documented Human Oversight & Guardrails and Security, Identity & Governance
- Lightpanda6.0 / 14Adds documented Human Oversight & Guardrails
- Modal8.0 / 14Adds documented Security, Identity & Governance and Memory & State Persistence
- Vocode6.0 / 14Adds documented Memory & State Persistence
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded