frp

frp is self-hosted reverse-proxy software for making intranet services reachable through a public-IP node. It is best suited to operators who can manage client and server binaries and want flexible tunneling without a subscription. Its protocol range and monitoring options are strong; XTCP’s dependence on NAT behavior and the Dashboard’s data-volume limit call for care in production.

Overview

frp handles TCP, UDP, HTTP and HTTPS proxying, with QUIC, KCP and WebSocket among the protocols available for client-server communication. That breadth suits varied services, from SSH and websites to DNS forwarding, Unix domain sockets, file serving and virtual network connections. Proxy groups add load balancing, while TCP and HTTP health checks can remove failed services from frps. These are useful controls for service operators, though frp remains infrastructure to run and maintain rather than a hosted tunnel service.

XTCP can use P2P hole punching so a successful connection sends traffic directly instead of through frps. That can avoid relaying traffic, but connection success and stability depend on the networks’ NAT types. The project is under development, and version 2 will not be compatible with version 1, so teams planning upgrades should account for a breaking transition.

Key features

Authentication and configuration

Client and server authentication supports tokens and OIDC, with tokens used by default. The project also includes encryption and documents communication security features. Configuration files can use TOML, YAML or JSON; INI remains supported but is no longer recommended. The choice of familiar formats helps fit frp into different operating practices, while the self-hosted setup still puts configuration and operations in the user’s hands.

Monitoring and administration

frps provides in-memory monitoring and Prometheus monitoring, and its browser Dashboard shows status and proxy statistics. Prometheus requires the Dashboard to be enabled. The Dashboard is not optimized for large volumes of proxy data, and the documentation advises against enabling it when access is slow; operators with substantial proxy data should weigh that constraint before relying on it. frpc has a separate Admin UI for proxy status and configuration.

Deployment

Written in Go, frp provides platform-specific binaries for cross-platform deployment without additional dependencies. This keeps deployment self-contained across supported systems, but users still need to run both the client and server in their own environment.

Pricing

frp costs nothing. The free plan is open-source reverse-proxy software with self-hosted client and server binaries, and the pricing model is free. It fits individuals and organizations prepared to operate their own infrastructure; there is no lower paid tier to trade off against, but self-hosting is part of the offer rather than an optional convenience.

Platforms

frp supports Android, Linux, macOS and Windows, alongside self-hosted deployment. Its configuration method is both file-based and through a web interface, and WebSocket support is included. Rate limiting is also supported.

Who it's for

Choose frp when you need to expose intranet services through a public-IP node and want options spanning common proxy protocols, authentication methods, health checks and monitoring. It is a practical fit for technically capable operators who value self-hosting and can account for version changes. Look elsewhere if your use depends on predictable XTCP connections across varying NAT types or a Dashboard that can handle large amounts of proxy data.

Pros and cons

Pros

• Broad support for TCP, UDP, HTTP and HTTPS, plus QUIC, KCP and WebSocket communication, serves more than web-only tunneling needs.
• Token and OIDC authentication, load balancing and TCP or HTTP health checks provide useful controls for managing access and service availability.
• Free, self-hosted binaries and cross-platform deployment without additional dependencies avoid subscription costs and extra runtime dependencies.
• In-memory and Prometheus monitoring, plus separate browser interfaces for frps and frpc, give operators visibility into proxy status and statistics.

Cons

• XTCP depends on NAT hole-punching success, so direct P2P connections cannot be counted on across different network types.
• The Dashboard is not suited to large volumes of proxy data and may be a poor fit when it loads slowly.
• Version 2 will not be compatible with version 1, complicating continuity for existing deployments.
• Self-hosting requires operators to manage the client, server and supporting infrastructure themselves.

Alternatives

Compare reverse proxy software if you want to consider a broader set of tools.

• Fabio is another free, self-hosted option if Consul-based service discovery fits your setup; its open-source software is MIT-licensed and requires Consul for discovery.
• freenginx is a free alternative for Linux, macOS, Windows or self-hosted use, with sources and documentation under a 2-clause BSD-like license.
• LocalXpose may suit readers who prefer a freemium tunnel service: its free Starter plan allows two active HTTP or HTTPS tunnels, SSL/TLS encryption and unique subdomains, with time limits and an interstitial warning page.
• ngrok is worth considering for a freemium option with a free trial; its free plan includes one-time usage of $5, up to three online endpoints and up to 1GB of data transfer.
• Openport offers an enterprise plan billed per month and device, with a minimum charge for 30 devices; consider it if that per-device model suits your deployment.
• Pomerium is a freemium alternative whose Personal Free plan includes 10 users, 10 routes, five policies and two service accounts.
• Zoraxy is another free, self-hosted choice, intended to be free of charge and released as AGPL open source.
• Envoy Proxy is a free open-source proxy available via Docker, macOS Homebrew or a static binary.

Verdict

frp is a strong choice for technically capable users who want free, self-hosted proxying across a wide range of protocols, with authentication, health checks and monitoring built in. Its main trade-offs are operational ownership, uncertain XTCP behavior across NAT types, a Dashboard unsuited to heavy proxy data and an incompatible version 2 transition. Choose it when those limits fit your environment; look elsewhere when dependable P2P behavior or large-scale Dashboard monitoring is central.