Deploy and Host GPT Researcher on Railway

GPT Researcher is an open-source AI-powered deep research agent that automates comprehensive web and document research. It combines planning, web search, crawling, retrieval, and report generation to produce detailed, factual, and citation-backed research reports. With support for multiple AI providers, local documents, and customizable workflows, GPT Researcher is ideal for developers, researchers, and organizations seeking reliable AI-assisted research.

About Hosting GPT Researcher

Hosting GPT Researcher on Railway provides a production-ready environment for running an AI-powered research platform without managing infrastructure. Railway automatically builds the application using the included Dockerfile, provisions HTTPS, manages deployments, and scales your service as needed. Configure your preferred AI provider using environment variables, optionally attach a persistent volume for local document research, and deploy a fully functional research application in minutes. GPT Researcher supports multiple AI providers, web search, local document analysis, LangSmith tracing, and AI-generated images, making Railway an ideal platform for self-hosting.

Common Use Cases

Generate comprehensive research reports using AI-powered web search with citations.
Analyze and summarize local documents such as PDFs, Word files, PowerPoint presentations, Markdown, CSV, and Excel files.
Build custom AI research assistants, knowledge management systems, or enterprise research workflows powered by OpenAI-compatible models.

Dependencies for GPT Researcher Hosting

OpenAI-compatible API provider (OpenAI, OpenRouter, Azure OpenAI, or other compatible providers).
Tavily Search API (recommended) for high-quality web research results.

Deployment Dependencies

GitHub Repository: https://github.com/assafelovic/gpt-researcher
Documentation: https://docs.gptr.dev
Tavily API: https://tavily.com
OpenAI API: https://platform.openai.com
LangSmith (Optional): https://smith.langchain.com

Implementation Details

Railway automatically detects the included Dockerfile and builds the application—no custom build or start commands are required.

Public Networking

Configure the Railway public endpoint as follows:

Setting	Value
Proxy Type	HTTP Proxy
Target Port	8000

Required Environment Variables

Add the following variables from your Railway service Variables tab:

Variable	Required	Description
`OPENAI_API_KEY`	✅	API key for your OpenAI-compatible provider.
`TAVILY_API_KEY`	✅	Tavily Search API key used for web research.

Example:

OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxx
TAVILY_API_KEY=tvly-xxxxxxxxxxxxxxxx

Optional Environment Variables

Variable	Description
`OPENAI_BASE_URL`	Custom OpenAI-compatible API endpoint (OpenRouter, Azure OpenAI, etc.).
`LANGCHAIN_TRACING_V2`	Enable LangSmith tracing (`true`).
`LANGCHAIN_API_KEY`	LangSmith API key.
`LANGCHAIN_PROJECT`	LangSmith project name.
`GOOGLE_API_KEY`	Required for AI-generated inline images.
`IMAGE_GENERATION_ENABLED`	Enable inline image generation (`true`).
`IMAGE_GENERATION_MODEL`	Example: `models/gemini-2.5-flash-image`.

Optional Persistent Storage

If you want GPT Researcher to analyze local documents:

Open your Railway service.
Navigate to Settings → Volumes.
Click Add Volume.
Configure the mount path as:

/app/my-docs

Then add the following environment variable:

DOC_PATH=/app/my-docs

GPT Researcher will automatically use this directory for supported local documents, including PDF, DOCX, PPTX, CSV, Excel, Markdown, and text files.

Deployment Notes

Railway automatically builds the project using the included Dockerfile.
No custom build command is required.
No custom start command is required.
Railway automatically injects the PORT environment variable, and the application starts on port 8000.
After deployment, generate a Railway public domain from Settings → Networking → Generate Domain to access the application.

Why Deploy GPT Researcher on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying GPT Researcher on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.