This document covers three standardized methods to acquire required configuration files for self-hosted FastGPT via Docker, for engineering and technical deployment workflows.
AI Agent-Assisted Configuration
Automate deployment setup by copying the following reference prompt to your coding agent:
Refer to https://doc.fastgpt.cn/deploy/SKILL.md and deploy FastGPT with Docker for me.Interactive Script Deployment
Run the base guided deployment command in Linux, macOS, or Windows WSL:
FASTGPT_DEPLOY_BASE_URL=https://doc.fastgpt.cn bash <(curl -fsSL https://doc.fastgpt.cn/deploy/install.sh)For deployments using custom domains, internal networks, or local addresses, set FASTGPT_DEPLOY_BASE_URL to your FastGPT documentation site root or a URL ending in /deploy to select the correct download source.
Non-interactive mode requires these mandatory environment variables:
| Variable | Description | Example |
|---|---|---|
FASTGPT_FE_DOMAIN | Full public access URL for FastGPT | https://fastgpt.example.com |
FASTGPT_SANDBOX_PROXY_URL | Sandbox WebSocket access URL | wss://sandbox-proxy.example.com |
FASTGPT_SANDBOX_PREVIEW_PROXY_URL | (Version 4.16+) HTTP preview proxy URL | Prompted interactively for older versions |
Interactive mode will automatically prompt for required values based on your target FastGPT version: version 4.15 only asks for the WebSocket URL, while 4.16 adds the HTTP preview URL prompt.
The script automates core deployment setup tasks:
- Downloads the official
docker-compose.ymlfile - Configures externally accessible S3 and MCP addresses
- Generates secure random credentials including the
rootlogin password, service tokens, app keys, and component passwords - Detects and updates the host Docker socket mount path in the compose file
After execution, the terminal will print the generated root password. Retain the final docker-compose.yml file securely, as it contains all generated credentials and is required for future upgrades. To use a local existing compose file, either select the local file option during the script prompt or pass the path via:
FASTGPT_LOCAL_COMPOSE_PATH=/path/to/docker-compose.yml bash <(curl -fsSL https://doc.fastgpt.cn/deploy/install.sh)Manual Configuration Download
For pinned deployments to specific docker-compose.yml versions, follow this step-by-step workflow:
- Download the targeted compose file for your database and region. Example command for Pgvector on the China mirror: ``
bash curl -fsSL https://doc.fastgpt.cn/deploy/docker/v4.15/cn/docker-compose.pg.yml -o docker-compose.source.yml`` Available database and region combinations are listed in the official documentation dropdown. - Download the installation script to your server: ``
bash curl -fsSL https://doc.fastgpt.cn/deploy/install.sh -o install.sh`` - Run the script with your local source compose file to generate the final deployment configuration: ``
bash FASTGPT_LOCAL_COMPOSE_PATH=./docker-compose.source.yml bash install.sh``
The script will copy the source compose file to docker-compose.yml, generate required credentials, and configure S3/MCP addresses as needed.
Source: FastGPT official source
Applicability and version scope
Use this page for the documented Deployment and upgrades scenario. Confirm the FastGPT, dependency, API, and deployment versions in the official source before applying a change.
Safety guardrails
Use [REDACTED_CREDENTIAL] for credentials and private data. Confirm the documented environment and version before review.
Rollback guidance
Restore the prior technical-content authority snapshot. Restore saved configuration and data snapshots, then repeat the smallest verification scenario.