Overview
ClawX is a desktop application that puts a graphical interface on OpenClaw, turning command-line agent orchestration into something you can install and click through. It runs on macOS, Windows and Linux as an Electron app with a React front end, and it embeds the OpenClaw core rather than asking you to install and manage the runtime separately — the project tracks upstream closely so it stays compatible with the wider OpenClaw ecosystem.
The problems it targets are setup and operations rather than model capability: a guided wizard replaces CLI setup, a unified provider panel replaces environment-variable hunting and YAML, Gateway process lifecycle is managed automatically, and skills are installed locally with an optional extension-provided marketplace. Startup update checks prompt before downloading or installing a new version.
Day to day it is a chat client with agent machinery behind it: multi-session history, streaming Markdown with syntax highlighting, KaTeX and tables, direct `@agent` routing, inline `/skill` cards, embedded subagent status with read-only drill-down into child runs, and read-only previews for Markdown, `.docx`, `.pptx` and local HTML. Scheduled tasks and channels are managed from the same UI. Two features sit behind Developer Mode: voice dictation, which transcribes through a speech-to-text service you configure, and local Computer Use, which is off by default and requires a vision-capable model.
What it does
- Embeds the official OpenClaw core, so there is no separate runtime installation to manage
- Setup wizard covering locale, AI providers (API key or OAuth device/browser login), skill bundles and a verification step
- Visual settings with real-time validation and a unified multi-provider configuration panel
- Automatic Gateway lifecycle management and startup update checks that prompt before installing
- Chat with multi-session history, streaming Markdown, `@agent` routing, `/skill` cards and live subagent drill-down
- Scheduled (cron) tasks and channel management from the desktop UI
- Local-first skill management, with an optional extension-provided marketplace
- Optional Computer Use, off by default and enabled through Developer Mode; needs a provider/model that accepts image input
- Context compaction that reserves 25% of a configured model context window (50,000-token fallback) and continues completed turns through the summary
Getting started
ClawX needs macOS 11+, Windows 10+ or Linux (Ubuntu 20.04+), 4GB RAM minimum (8GB recommended) and 1GB of disk. Local Computer Use additionally requires macOS 13+ or Windows 10+ on x64/arm64.
Install a pre-built release
The recommended path is to download the build for your platform from the project's GitHub Releases page. No terminal or separate OpenClaw install is required.
Or build from source
Clone the repository, initialize it and start in development mode.
git clone https://github.com/ValueCell-ai/ClawX.git
cd ClawX
pnpm run init
pnpm devWork through the setup wizard
On first launch the wizard walks through four steps: language and region (it preselects your system language when supported, English otherwise), AI provider credentials via API key or OAuth, a choice of pre-configured skill bundles, and a verification step that tests the configuration before you reach the main interface.
Turn on Developer Mode for the optional features
Voice dictation and Computer Use are both gated behind Developer Mode in Settings. For dictation, configure a speech-to-text service under Models → Speech-to-text, choosing the API type and a preset. For Computer Use, enable the feature in the sidebar page that appears; the choice persists across restarts, and you must select a provider and model that support image input or the agent will be typing blind.
Commands and code are distilled from the project's own documentation — always check the official repo for the latest.
When to use it
- Running OpenClaw agents without setting up a CLI, config files or environment variables
- Managing several model providers and switching between them from one settings panel
- Scheduling recurring agent work and watching subagent runs from a desktop UI rather than log output
- Giving non-terminal users on Windows, macOS or Linux access to an agent stack that is otherwise CLI-first
How ClawX compares
ClawX alongside other open-source assistants & chatbots tools AI/TLDR tracks, ranked by GitHub stars.
| Tool | Stars | What it does |
|---|---|---|
| OpenClaw | ★ 390k | OpenClaw is a self-hosted personal AI assistant that answers you on WhatsApp, Telegram, Slack, Discord, and many other channels, with voice and a live visual canvas. |
| Hermes Agent | ★ 247k | A self-improving personal AI agent from Nous Research that builds skills from experience, remembers across sessions, and reaches you on Telegram, Discord, Slack, and more. |
| Odysseus | ★ 87.4k | A self-hosted AI workspace that puts chat, agents, deep research, documents, email, notes, tasks and calendar behind one Docker Compose stack, over local or API models. |
| CowAgent | ★ 47k | A self-hosted assistant that plans and executes tasks with built-in file, terminal, browser and search tools, and answers across a web console plus a dozen messaging platforms. |
| AstrBot | ★ 40.6k | An all-in-one agent chatbot platform that puts LLM conversations, tools, knowledge bases and a plugin marketplace inside messaging apps like Telegram, Slack, Discord, QQ and Feishu. |
| OpenHuman | ★ 39.9k | A local-first desktop personal AI for macOS, Windows and Linux that keeps a compressed memory tree on your machine and orchestrates checkpointed research and automation workflows. |
| MindsHub | ★ 39.7k | An agent workspace for knowledge work and software development that runs swappable open-source agent harnesses against your choice of frontier or open models. |
| ClawX | ★ 7.6k | Desktop app that wraps the OpenClaw agent runtime in a graphical interface, with the runtime embedded |