Backed by 56,000+ GitHub stars and over 240 million Docker pulls, LiteLLM delivers the open-source AI gateway trusted by Netflix, Lemonade, Rocket Money, and thousands of engineering teams to route every LLM request through one unified API. The Rust-core gateway adds sub-millisecond overhead per request with 8ms P95 latency at 1,000 RPS, 15x throughput improvement and 11x lower memory footprint compared to Python-only proxies. A single OpenAI-compatible endpoint connects to 100+ providers and 1,800+ models spanning OpenAI, Anthropic, Google Gemini, AWS Bedrock, Azure OpenAI, Vertex AI, Hugging Face, vLLM, Nvidia NIM, Ollama, and Mistral with day-zero support for new model releases. The Auto Router V2 classifies request complexity across four tiers using rule-based scoring, semantic keyword matching, and adaptive Thompson sampling to route each request to the most cost-effective model without API calls or training data. Virtual API keys enable multi-tenant governance with per-team, per-user, and per-project cost tracking, budget caps with automatic fallback rerouting, and role-based access control. Built-in guardrails provide PII masking, prompt injection detection, and model-graded evaluation before requests reach providers. The Agent Gateway extends routing from model calls to agent workflows with MCP server integration. Observability integrates with Langfuse, Arize Phoenix, OpenTelemetry, and MLflow for complete request tracing. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Your apps keep using the OpenAI SDK while Antirion, a self-hosted AI gateway, decides which provider answers, what each team may spend, and what gets logged. Clients speak the OpenAI, Anthropic, Gemini, or Bedrock APIs, and Antirion translates between them, so a Claude deployment can answer an OpenAI-style call, streams included. Add providers such as Amazon Bedrock, Vertex AI, Mistral, Groq, DeepSeek, or a vLLM server, then publish them as named Models. Routing rules send traffic by team or project with fallback chains, weighted splits, least-cost or least-latency targets, retries, hedging, and circuit breakers. Budgets and rate limits in dollars or tokens apply per organization, team, project, and key, with projections and alerts before money runs out. Guardrails check for personal data, secrets, keywords, and regex matches, then block, redact, flag, or divert. Exact and semantic caching answers repeated prompts without calling a provider. The request log records every attempt, route, token count, cost, and guardrail verdict, with Prometheus metrics and OTLP traces. An MCP gateway and A2A proxy put tools and agents behind the same keys. Sign in with OIDC, SAML, or GitHub, provision users with SCIM, and review every change in the audit log. Changes you choose can require a change request or a second person's approval, and the Playground sends requests from the console. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

BitRouter is a context-aware LLM router that learns which model delivers the cheapest successful outcome per workflow step, cutting agent costs by up to 80% while maintaining 96% quality versus all-frontier baselines. Point any agent runtime at http://localhost:4356 with a one-line OPENAI_BASE_URL change and BitRouter routes to OpenAI, Anthropic, Google, Groq, DeepSeek, Mistral, Moonshot, MiniMax, Nvidia, and any OpenAI-compatible endpoint simultaneously, normalizing authentication, streaming, and cross-protocol translation between wire formats. The act-observe-evaluate-learn loop traces every hop with cost, tokens, and latency attribution, scores each decision against a versioned policy-lock.yaml, then tightens routes automatically with no LLM judge in the path. Native MCP gateway auto-discovers tools from connected servers and makes them routable and governed alongside model calls. Agent Client Protocol integration enables the TUI to manage Claude Code, Codex, OpenCode, OpenClaw, Gemini, and Copilot sessions in real time with inline tool-call approval and live streaming. Built-in guardrails inspect, redact, or block risky content before requests leave your network. Virtual keys scope API access per agent or user without exposing upstream credentials. Per-agent spend caps and loop guards contain runaway cost automatically. Multi-account failover reroutes mid-run so rate limits never re-pay completed work. Ships as a single Rust binary via npm or Cargo. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

Forge intercepts failing LLM tool calls and fixes them before they derail your agent workflow, applying rescue parsing, retry nudges, response validation, and step enforcement between your AI clients and local model backends. The proxy server mode drops in as a transparent intermediary speaking both the OpenAI chat-completions API and the Anthropic Messages API, so tools like Aider, Claude Code, Continue, and opencode connect through it without configuration changes. Under the hood, the WorkflowRunner provides a complete agentic loop manager with system prompt injection, tool execution, context compaction with configurable thresholds, and VRAM budgeting for consumer GPUs with 12-32 GB. SlotWorker enables priority-queued access to shared inference slots with automatic preemption for multi-agent architectures. The guardrails middleware exposes a two-method check-and-record API that wraps into any existing orchestration loop, providing malformed tool-call rescue parsing, retry nudge generation, required step enforcement, and prerequisite ordering without taking over execution control. Backend adapters support generic OpenAI-compatible endpoints, Ollama, llama-server, Llamafile, vLLM, and Anthropic with automatic model discovery and health checking. Architecture Decision Records document every design choice. Launched February 2026, already at 2,200+ GitHub stars. MIT licensed.

Providing an intuitive control plane for modern reverse proxy infrastructure, CaddyUI enables system administrators to manage proxy hosts, automated TLS certificates, and traffic routing without hand-editing configuration files. Operators can create and adjust reverse proxy destinations, URL path rewrites, custom headers, and redirection rules using structured forms that validate against the proxy engine before saving. The platform coordinates automatic ACME certificate requests and renewals, displaying live certificate expiration timelines, validation stages, and internal self-signed certificate authority options for local network services. Fleet management tools synchronize proxy rules across multiple distributed server instances, preventing configuration drift across staging and production clusters. Automated post-deployment health checks verify upstream endpoint availability, HTTP status codes, and latency thresholds, instantly rolling back to the previous configuration if an expectation fails. An integrated observability suite delivers real-time visitor traffic analytics, geographic maps, HTTP response code breakdowns, and single-click client IP blocking to counter malicious probing attempts. Administrators can schedule automated database backups, configure multi-user access with role permissions, and integrate CrowdSec threat intelligence to protect backends. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

Anyone who pays for Claude Code, Codex, and Gemini CLI separately ends up juggling three endpoints, three sets of keys, and three kinds of rate-limit surprise, and CliRelay collapses all of that into one self-hosted gateway with a control panel. Clients speak the OpenAI Chat or Responses, Anthropic Messages, or Gemini protocol, and CliRelay translates each request for whichever upstream account is healthy. Channel groups limit where each key may send traffic, round-robin, fill-first, and session-sticky scheduling spread it, and cooldowns with automatic failover move requests off accounts that have run out of quota. Add upstreams from the /manage console through browser sign-in, device codes, or bulk credential import, covering Claude, Codex, Gemini, Antigravity, Vertex AI, AWS Bedrock, Grok, Qwen, Kimi, and any OpenAI-compatible service. Portal accounts own client keys with daily, period, and lifetime quotas, RPM and TPM limits, and per-model pricing. A monitor center scores health and charts live traffic, P50 to P99 latency, and time to first token, while every request lands in a searchable log with tokens, latency, and cost. Tenants, roles with resource.action permissions, content moderation profiles, an outbound proxy pool, and an audit trail make shared deployments workable, and the console speaks English, Simplified Chinese, and Russian. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Created to eliminate the configuration fragmentation plaguing multi-device Homelab setups and small-team infrastructures, NetsGo merges what typically requires three separate tools — a tunnel daemon, a management panel, and a monitoring stack — into a single Go binary that starts with one command and listens on one port. The server exposes HTTP, TCP, and UDP tunnels, a built-in Web console for node management, a REST API for automation, and real-time traffic statistics with per-client bandwidth limiting, all sharing default port 9527 so firewalls and reverse proxies need only a single rule. Version 0.1.11 introduced Client-to-Client P2P tunnels via WebRTC DataChannel with three selectable transport strategies — Server relay, Prefer P2P with automatic fallback, and P2P only — reducing latency for direct connections while maintaining the centralized control plane. The management console displays every connected client's online status, authentication tokens, tunnel configuration, active transport mode, P2P candidate gathering progress, and cumulative byte-level traffic counters updated every second. Unlike frp or rathole which delegate management to external plugins or configuration files, NetsGo provides built-in client key generation, tunnel target migration between nodes, SOCKS5 proxying, IP-whitelist console access, and one-command server upgrades. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

Your scanner still emails PDFs, your backup scripts still send nightly reports, and none of them will ever learn OAuth 2.0: Email OAuth 2.0 Proxy learns it on their behalf. It listens on local IMAP, POP, and SMTP ports on your server, and when a client logs in the old-fashioned way with a username and password, the proxy swaps that exchange for a modern SASL XOAUTH2 token before the connection ever reaches the provider. The client never knows anything changed, so ancient mail apps, embedded devices, monitoring tools, and hand-rolled scripts keep sending and fetching mail through Outlook, Office 365, Gmail, Google Workspace, Fastmail, Yahoo, AOL, and any other provider that demands OAuth 2.0. The hop from your client to the proxy stays simple, while the proxy opens a fully encrypted TLS connection upstream, including STARTTLS negotiation for SMTP relays that require it. One proxy instance serves many accounts at once, with each mailbox mapped to its own local port and its tokens cached under a password you control. Optional plugins can even rewrite traffic in flight: strip Microsoft Defender Safe Links back to their real URLs, stop duplicate sent messages, or swallow test mail into a black hole. Setup is a single configuration file covering server ports, per-account scopes, and token endpoints, and the proxy then runs quietly as a headless service. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Where Traefik's built-in dashboard only shows what is running, Traefik Manager lets you change it: add routes, middlewares, and services from forms and wizards instead of editing YAML by hand. A first-run wizard connects to the Traefik API, creates a route for the manager itself, and sets the admin password. After that, the dashboard turns every route into a launchable app card with live health, and the Route Map draws each request from entry point through middlewares to backend. New routes use a slide-in form for HTTP, TCP, or UDP with several domains and backends, sticky sessions, and health checks. 30 middleware wizards handle authentication, rate limiting, headers, CORS, and redirects, and weighted, mirroring, or failover services can be built on their own. For day-to-day monitoring, the Certificates tab tracks expiry from acme.json, the Logs tab turns the access log into clickable analytics, the CrowdSec tab lists attacks and bans, and alerts reach Discord, Slack, ntfy, Telegram, and more. Changes to traefik.yml apply with a one-click restart, every edit is backed up locally and can be pushed to a git repository, and a lightweight agent brings remote Traefik servers into the same interface. Sign in with TOTP two-factor or OIDC single sign-on, in any of 14 languages. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. GPL-3.0 licensed.

Inference Gateway puts a single OpenAI-compatible API endpoint in front of OpenAI, Anthropic, Groq, Cohere, Ollama, DeepSeek, Google, Mistral, MiniMax, Moonshot, Nvidia, and llama.cpp, so your application code never changes when you switch models or providers. The Go binary starts on port 8080 and normalizes authentication, streaming protocols, and response formats across all backends transparently. Native Model Context Protocol support auto-discovers tools from connected MCP servers and injects them into LLM requests without client-side management, enabling server-side tool execution across any provider that supports function calling. Agent-to-Agent protocol integration allows distributed agent communication through a declarative Agent Definition Language that generates production-ready Go or Rust servers from a single YAML manifest. The dedicated Kubernetes Operator manages Gateway, Agent, MCP, and Orchestrator custom resources with automatic HPA scaling, OIDC authentication, and service discovery that rebuilds MCP configurations when the discovered server set changes. Prometheus metrics and OpenTelemetry tracing provide full request-level observability across the entire inference pipeline. Middleware controls enable per-request provider selection, model routing, and fallback strategies. Official SDKs in Go, Python, TypeScript, and Rust provide typed client interfaces with streaming support. Docker Compose deployment requires only environment variables for API keys. A CNCF Sandbox applicant. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

Tag a Docker container with two labels, and AutoXpose turns it into a published service with a DNS record, a reverse proxy route, and HTTPS, without you opening a single provider dashboard. From the web dashboard you connect one of six DNS providers, Cloudflare, Netlify, DigitalOcean, Porkbun, Aliyun, or Tencent DNSPod, and point it at the Nginx Proxy Manager or Caddy instance you already run. It watches your containers continuously after that: a new label creates the record, propagation is verified before the service shows as live, and removing the container removes its public entries too. Auto-expose mode publishes containers the moment they appear, while manual mode stages them in the dashboard for approval first. Color-coded service tags sort what you run into web, database, and media groups at a glance, and dashboard warnings flag misconfigurations before they become broken links. Nginx Proxy Manager access lists attach by label, so an admin panel stays reachable only from the networks you trust. When containers vanish, orphaned cleanup hunts down the stale DNS records and proxy hosts they left behind. Settings and state live inside the same single container, so the publishing pipeline for a homelab or a small fleet stays one compose service you can back up as a unit. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

A self-hosted front door for apps that live on home labs, office LANs, and private clusters, Wiredoor publishes them to the internet without opening a single inbound port on the private side. The Wiredoor CLI on each remote machine dials out to the server over an encrypted WireGuard tunnel. Web apps and APIs are routed by domain and path, WebSocket traffic included, while TCP and UDP services such as SSH or Redis get a public port of their own. Public domains receive Let's Encrypt certificates, internal names fall back to self-signed ones, and Cloudflare or GoDaddy credentials let the server create DNS records itself. Any web route can demand OAuth2 sign-in with an email allow list and a few open paths, and every protocol accepts IP allow lists. Gateway nodes stretch one tunnel across a Docker network, Kubernetes cluster, or private subnet, so a single link serves many backends. Services can be switched off or given an expiry such as 30 minutes, 2 hours, or 1 day. The dashboard shows which nodes are connected and whether their services are enabled, while the CLI creates, enables, and disables services from a terminal. Optional Prometheus and Grafana views chart traffic and tunnel peers, and CrowdSec can block abusive addresses. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Rustunnel turns a service running on your laptop or private network into a public HTTPS, TCP, or UDP endpoint without opening inbound ports on the machine where it lives. Point the client at a local port and share the public address for webhook testing, a quick demo, an AI agent tool, SSH access, a game server, or an IoT device that cannot accept inbound connections. Give tunnels custom subdomains on your own wildcard domain, terminate HTTPS with automatic Let's Encrypt certificates, and keep client traffic encrypted from edge to local service. Put several backends behind one subdomain or TCP port and let TCP or HTTP health probes remove sick members before visitors hit them. When two peers can reach each other directly, start a P2P tunnel with NAT hole punching and QUIC, with relay fallback when the direct path fails. Watch sessions, tunnel health, traffic, and request logs from the live dashboard, inspect headers and bodies in the local request inspector, and replay a captured call while debugging a webhook. Manage tunnels, tokens, groups, captured requests, and history through the REST API, scrape Prometheus metrics, review audit logs, or let an MCP-compatible coding assistant create and close tunnels in plain language. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Usulnet packs container management, Trivy security scanning, Nginx reverse proxy, scheduled backups, WireGuard VPN, firewall rules, and multi-node orchestration into a single 70MB Go binary with zero external runtime dependencies. Every module ships in one download with no paid tiers, no telemetry, and no edition gating. Container lifecycle management covers creation, start, stop, restart, pause, kill, and removal with bulk operations, real-time resource statistics, filesystem browsing, and settings editing. Trivy integration scans images and running containers for CVEs with severity scoring, generates SBOMs, and validates CIS benchmarks. The multi-node architecture supports standalone, master, or agent modes where agents connect over NATS JetStream with mTLS encryption, enabling remote Docker host management from a central dashboard. Reverse proxy configuration handles Nginx with automatic Let's Encrypt certificates, TCP/UDP stream proxying, access lists, and dead host detection. Backup operations capture container volumes and Compose stacks on configurable schedules with retention policies and full restore capabilities. A built-in application catalog provides 60+ one-click templates for common services. JWT license validation uses an RSA-4096 public key embedded in the binary, requiring no call-home and working entirely offline. For teams tired of maintaining separate tools for each infrastructure concern, Usulnet collapses the entire stack into a single point of management with Docker Compose deployment alongside PostgreSQL, Redis, and NATS.