1Panel
The only open-source VPS control panel with native AI agent runtime — deploy websites, Docker stacks, and local LLMs from one web interface.
1Panel is a modern, open-source VPS control panel that simplifies Linux server management through a clean web interface, letting you deploy websites, Docker containers, AI agents, and over 165 open-source applications without touching the command line. It is trusted by over two million self-hosters worldwide and is uniquely positioned as the only control panel with a built-in AI agent runtime, supporting Ollama models, OpenClaw personal agents, Hermes agents, vLLM inference, and TensorRT LLM deployments.
The panel is architecturally split into a central core control server and per-host agent daemons. The core handles authentication, settings, and the web UI, while agents run on each managed Linux server to orchestrate containers, websites, databases, file systems, and security policies. Both communicate over secured channels, allowing 1Panel to scale from a single VPS to multi-node fleets managed through one dashboard.
1Panel’s ecosystem includes automated SSL via Let’s Encrypt, firewall and WAF management, real-time system monitoring with GPU tracking, scheduled cloud backups to AWS S3 and Cloudflare R2, and a curated app marketplace with one-click installs and updates. It installs in about 60 seconds via a single bash script and runs as a lightweight systemd service on Debian, Ubuntu, CentOS, Rocky Linux, and compatible distributions.
What You Get
- Native AI Agent Runtime - Deploy Ollama models, OpenClaw personal agents, Hermes conversational agents, and vLLM or TensorRT LLM endpoints directly from the dashboard with GPU monitoring and per-agent model account configuration.
- One-Click Website Deployment - Launch production-ready websites with automatic domain binding, Let’s Encrypt SSL provisioning, OpenResty/Nginx configuration, reverse proxies, and redirect rules — zero manual setup required.
- App Marketplace with 165+ Apps - Browse and install trusted open-source tools (Nextcloud, Bitwarden, NocoBase, Umami, phpMyAdmin, JumpServer) with a single click, complete with automatic dependency resolution and upgrade management.
- Docker and Container Management - Create, start, stop, and inspect containers, images, networks, volumes, and Compose orchestrations through a visual UI with log viewer, port mapping controls, and bulk image import.
- Automated Backup and Restore - Schedule encrypted backups to local storage, AWS S3, Cloudflare R2, or OneDrive with configurable retention policies and one-click snapshot restore, including cross-platform recovery.
- Real-Time System and GPU Monitoring - Track CPU, memory, disk, and network usage in real time with historical graphs, configurable alert thresholds, and dedicated GPU utilization panels for AI workloads.
- Web-Based File Manager - Browse, upload, edit, compress, and manage file permissions directly in the browser, with file remarks, last-opened-file memory, and FTP integration support.
- Database Visual Management - Manage MySQL, MariaDB, PostgreSQL, MongoDB, and Redis instances via a GUI with query execution, import/export, user management, and one-click full connection URL copy.
- Firewall, WAF, and Audit Logging - Configure iptables-based firewall rules, WAF blocking, fail2ban, and view detailed audit logs of user actions and system events for security compliance.
- Multi-Node Fleet Management - Connect and manage multiple Linux servers from a single dashboard with node overview pages, per-node application deployments, and task manager visibility across all hosts.
Common Use Cases
- Self-hosting a personal cloud stack - A developer uses 1Panel to deploy Nextcloud for file sync, Bitwarden for passwords, and Umami for analytics on a single $5/month VPS, replacing multiple SaaS subscriptions without writing Nginx configs.
- Running local AI models on a VPS - A team deploys Ollama with Qwen or DeepSeek via 1Panel’s AI agent runtime to power internal chatbots and coding assistants, with GPU utilization monitoring and per-model API account management.
- Managing client servers as an MSP - A managed service provider uses 1Panel’s multi-node dashboard, audit logs, firewall controls, and role-based user accounts to securely administer dozens of VPS instances from one interface.
- Deploying a WordPress or PHP site - A web agency installs WordPress or a custom PHP runtime with one click, configures automatic SSL renewal, sets up a subdomain reverse proxy, and enables website uptime monitoring — all within five minutes.
- Running AI coding agents for a development team - An engineering team deploys Hermes agents connected to GitHub repositories via 1Panel, letting developers trigger AI-powered code reviews and PR assistance directly from a managed web interface.
- Building a self-hosted SaaS alternative stack - A startup installs Discourse for community, NocoDB for databases, and Gitea for code hosting through 1Panel’s marketplace, maintaining full data ownership with automated daily backups to S3.
Under The Hood
Architecture
1Panel uses a hub-and-spoke architecture built from two independently compiled Go modules: core, which runs the central web server and control plane handling authentication, user settings, and the Vue 3 frontend, and agent, which runs as a daemon on each managed Linux server and owns all operational responsibilities — container orchestration, website management, database administration, AI agent lifecycle, and file system access. Both expose Gin REST APIs versioned under /api/v2, use the same structural conventions, and coordinate over authenticated HTTP channels. The agent module is substantially larger, with over 100 service files covering every managed subsystem. Within each module, a strict three-layer pattern is enforced: Gin handlers delegate to interface-defined service structs, which use repository packages backed by GORM models — ensuring no data access logic leaks into HTTP handlers or business logic. Enterprise-gated features are toggled at compile time using the xpack build tag system, which swaps stub implementations for licensed code without polluting the community codebase with conditional branches.
Tech Stack 1Panel’s backend is written in Go compiled to portable static binaries, with the Vue 3 frontend bundled as embedded assets using Vite and included in the binary at build time. Database persistence uses a pure-Go SQLite driver (glebarez/sqlite, CGO-free) for the panel’s own metadata, while managed databases — MySQL, MariaDB, PostgreSQL, MongoDB, Redis — are orchestrated as Docker containers. Container management integrates directly with the Docker daemon via the official Docker Go SDK, with Compose orchestrations handled through compose-spec/compose-go. Web server automation targets OpenResty (Nginx with Lua extensions), configured programmatically by the agent service layer. The AI runtime supports Ollama (HTTP API), vLLM (OpenAI-compatible inference), TensorRT LLM, and the proprietary OpenClaw and Hermes agent frameworks, each with dedicated service files and provider abstractions. WebAuthn and passkey-based authentication are implemented via go-webauthn for hardware key and biometric login. Multi-architecture release builds are automated with GoReleaser targeting AMD64, ARM64, RISC-V, and PPC64LE Linux targets.
Code Quality
1Panel demonstrates strong structural discipline across both modules. Services are always defined as Go interfaces, allowing clean dependency inversion and making the xpack stub pattern reliable. DTOs and request/response structs are validated with go-playground/validator, and a dedicated buserr package provides structured business errors with i18n message keys, preventing swallowed errors across API boundaries. The codebase has comprehensive Swagger documentation generated from source comments. Test coverage is limited — only a swagger generation test was found in the repository — so correctness relies on manual QA and the integration testing implied by active bug-fix release cadence (over 110 releases). Naming conventions are consistent and idiomatic Go throughout, and the Makefile and GoReleaser config reflect mature build automation.
What Makes It Unique 1Panel is the only open-source VPS control panel with a native, first-class AI agent runtime. Rather than treating AI as an app-store extension, the agent service layer has dedicated subsystems for Ollama model management, OpenClaw agent lifecycle, Hermes conversational agent sessions, vLLM inference endpoint management, and TensorRT LLM deployment — each with their own provider abstractions, channel configurations (Telegram, Discord, WeChat, DingTalk, Feishu), and skill systems. The xpack build tag architecture enables a clean OSS/Pro split without forking the codebase, with community builds receiving stubs for multi-node and advanced auth features. The cmux-based server allows HTTP and HTTPS traffic to coexist on a single port in Mux mode, reducing firewall complexity for self-hosters. Passkey and WebAuthn authentication are deeply integrated at the core auth layer rather than bolted on, providing phishing-resistant login for production deployments.
Self-Hosting
1Panel is released under the GNU General Public License v3.0 (GPL-3.0). For self-hosters, this means you can run, modify, and distribute the software freely, including for commercial purposes on your own servers. The copyleft clause applies if you distribute modified versions of 1Panel itself — if you run it internally or offer it as a hosted service to others, you are not required to open-source your infrastructure or applications running on top of it. The GPL-3.0 license is one of the stronger copyleft licenses, so teams embedding 1Panel into a redistributed product should review the terms carefully, but for the overwhelming majority of self-hosting use cases it imposes no practical restrictions.
Running 1Panel yourself requires a Linux VPS or dedicated server with at least 1 GB of RAM, Docker, and internet access during installation. The install script handles service setup, systemd registration, and initial configuration in about 60 seconds. Ongoing operations include managing your own uptime, applying 1Panel updates (which are released frequently — roughly bi-weekly), monitoring disk and memory growth as you add containers and backups, and maintaining SSL certificate renewals (automated via Let’s Encrypt). The agent-core architecture means each managed server runs its own agent process, so multi-server deployments multiply the operational surface. You are responsible for your own backup strategy, server hardening, and incident response — 1Panel provides the tools, but not the management layer.
The community (OSS) edition is free indefinitely and covers single-server management with one OpenClaw AI agent, basic WAF, Docker management, the full app marketplace, SSL, backups, and monitoring. The Pro edition, starting at $80 per year, adds unlimited AI agents, advanced WAF and website tamper protection, website uptime monitoring, multi-node fleet management, custom branding, and priority support. For teams managing more than one server or deploying AI agents at scale, the Pro tier addresses real operational gaps that the OSS edition does not cover. There is a 30-day free trial available for Pro.
Related Apps
Uptime Kuma
Monitoring
Self-hosted monitoring for every service you run — 23 monitor types, 95 notification channels, live dashboards, and public status pages with no vendor lock-in.
Uptime Kuma
MITNetdata
Monitoring · Devops
Real-time per-second metrics, ML-powered anomaly detection, and zero-config observability for any infrastructure.
Netdata
GPL 3.0Grafana
Monitoring · Analytics
The open-source observability platform that unifies metrics, logs, and traces from any data source into dynamic, queryable dashboards.