Skip to content

OpenClaw or RemoteClaw?

OpenClaw or RemoteClaw?

RemoteClaw is a fork of OpenClaw that replaced the embedded AI execution engine with a middleware architecture. Both projects are actively maintained but serve different use cases.

Quick decision

Already using Claude Code, Gemini CLI, Codex CLI, or OpenCode? Use RemoteClaw. It connects your existing agent to messaging channels without adding another LLM layer.

Want a self-contained platform with built-in model providers and a skills marketplace? Use OpenClaw. It includes everything in one process.

Decision tree

If you want…Choose
Connect an existing CLI agent (Claude, Gemini, Codex, OpenCode) to WhatsApp/Telegram/SlackRemoteClaw
Switch between multiple agent runtimes via configRemoteClaw
Middleware-only architecture (no embedded LLM overhead)RemoteClaw
MCP-first tool system (50 tools via MCP server)RemoteClaw
Subprocess isolation (agent crashes don’t crash the gateway)RemoteClaw
Consumer onboarding wizard with 30+ LLM provider pickerOpenClaw
Built-in skills marketplace (ClawHub)OpenClaw
In-process model catalog (switch between OpenAI, Anthropic, Google, etc.)OpenClaw
Docker sandbox for code executionOpenClaw
Centralized OAuth for all providersOpenClaw

Architecture comparison

RemoteClaw (middleware)
├── ChannelBridge (message routing + session tracking)
├── CLI subprocess (claude, gemini, codex, or opencode)
│ ├── LLM interaction (handled by the CLI)
│ ├── Tool execution (handled by the CLI)
│ └── Conversation management (handled by the CLI)
└── MCP server (injected into subprocess for gateway access)

The CLI agent owns the agentic loop. RemoteClaw handles session persistence, message delivery, and MCP tool bridging.

For a deeper architectural comparison, see Middleware Architecture — How This Differs from OpenClaw.

Switching from OpenClaw

If you decide to switch, see Migrate from OpenClaw for step-by-step instructions and Breaking changes for what to expect.