Deploy and Host QwenPaw on Railway

QwenPaw is a personal AI assistant from AgentScope. It runs in your own environment, has a web console, skills (documents, browser, scheduled tasks) and can connect to chat channels such as DingTalk, Feishu, QQ and Discord.

About Hosting QwenPaw

This template runs the official agentscope/qwenpaw:latest image on port 8088 with a persistent volume at /app/working (config, memory, skills and provider keys). Because the agent can run commands and read files, the template turns on QwenPaw's login: an admin username and a random password are generated for you and shown in the service's Variables tab (QWENPAW_AUTH_USERNAME, QWENPAW_AUTH_PASSWORD). Open the public URL, sign in with them, then add an LLM provider and its API key in the console settings (or use a local model endpoint).

Common Use Cases

A private AI assistant reachable from your browser and chat apps
Scheduled tasks and document or spreadsheet work through built-in skills
Connecting a model provider (DashScope, OpenAI and others) to your messaging channels

Dependencies for QwenPaw Hosting

One service with a volume; no database
An API key for the LLM provider you choose (entered in the console)

Deployment Dependencies

Why Deploy QwenPaw on Railway?

Railway gives you a public HTTPS URL, a persistent volume and automatic restarts with no server to maintain. The generated login keeps strangers out of an agent that can run tools, so keep the password private.