ServerKit is a lightweight, modern server control panel for managing web applications, databases, and services on your VPS or dedicated server. Built with Python/Flask backend and React frontend.
Python astracloudflarecontrol-panelcyberpaneldnsfablehosting-control-panelhosting-panelkimi-k3server-managementserverkitubuntu-web-panelwebserver
id: anythingllm
name: AnythingLLM
version: "1.0"
description: All-in-one private ChatGPT-style app with built-in RAG, AI agents, and document workspaces.
icon: https://serverkit.ai/imgs/template-icons/anythingllm.svg
website: https://anythingllm.com
documentation: https://docs.anythingllm.com
categories:
- ai
- llm
- chat
- rag
variables:
- name: HTTP_PORT
description: Port for the AnythingLLM web UI (auto-assigned)
type: port
default: "3001"
hidden: true
- name: JWT_SECRET
description: Secret used to sign AnythingLLM session tokens (auto-generated)
type: random
length: 48
hidden: true
- name: SIG_KEY
description: Signing key for AnythingLLM internal payloads (auto-generated)
type: random
length: 48
hidden: true
- name: SIG_SALT
description: Signing salt for AnythingLLM internal payloads (auto-generated)
type: random
length: 48
hidden: true
ports:
- port: "${HTTP_PORT}"
protocol: http
description: AnythingLLM Web UI
requirements:
memory: 2GB
storage: 5GB
compose:
services:
anythingllm:
image: mintplexlabs/anythingllm:latest
container_name: ${APP_NAME}
restart: unless-stopped
cap_add:
- SYS_ADMIN
ports:
- "${HTTP_PORT}:3001"
environment:
STORAGE_DIR: "/app/server/storage"
JWT_SECRET: "${JWT_SECRET}"
SIG_KEY: "${SIG_KEY}"
SIG_SALT: "${SIG_SALT}"
DISABLE_TELEMETRY: "true"
volumes:
- anythingllm-storage:/app/server/storage
healthcheck:
test: ["CMD", "curl", "-fsS", "http://127.0.0.1:3001/api/ping"]
interval: 30s
timeout: 10s
retries: 5
start_period: 30s
volumes:
anythingllm-storage:
post_install: |
AnythingLLM is installed and running!
- Web UI: http://your-server:${HTTP_PORT}
- On first launch, create your admin account and pick an LLM provider.
AnythingLLM does not bundle an LLM. Point it at any provider from the
in-app setup wizard (Settings -> LLM Preference):
- Ollama: set the base URL to http://<your-server>:11434 (or the
docker service URL if Ollama runs in the same network).
- Prompture Hub / any OpenAI-compatible endpoint: choose the
"Generic OpenAI" provider and supply the base URL + API key.
Uploaded documents and the local vector store live in the
anythingllm-storage volume.