Runs on Your Machine
Mac, Windows, or Linux. Anthropic, OpenAI, or local models. Local files stay under your control; cloud model requests follow the selected provider.
OpeClaw is built for users who want models, files, prompts, and automation steps in one local workflow. Before installing, check the download page, current release status, and the permissions you are ready to grant.
Before installing, decide whether you actually need a local AI assistant workflow.
If you use AI for drafts, research notes, scripts, and repeatable checks every day, OpeClaw makes more sense as a workflow layer than as another chat box.
Local runtime does not mean every request stays offline. If you connect a cloud model, the prompt content for that request may be handled by the provider.
If you only ask occasional questions, a normal chatbot may be enough. If you need prompts, files, and automation in one loop, OpeClaw is worth tracking.
Check the download page before installing. Windows, macOS, and Linux setup entries should be verified instead of copied from reposted scripts.
For SEO writing, split the workflow into drafting, human editing, FAQ checks, and source review. That is where a personal AI assistant can save time.
For errors, record your system, model provider, network, and permissions first. Not every failure is an application bug.
Start with verifiable setup notes, model boundaries, and local permissions
Mac, Windows, or Linux. Anthropic, OpenAI, or local models. Local files stay under your control; cloud model requests follow the selected provider.
Usage notes focus on WhatsApp, Telegram, Discord, Slack, Signal, iMessage, and similar chat entry points. Availability depends on the current project state.
Organizes longer-term context around your preferences. What is retained depends on local configuration, model settings, and the workflows you authorize.
Useful for organizing web research, form steps, and repeated browser routines. Actual capability depends on current project notes and your authorization.
Designed for file, script, and local automation workflows. When shell or file permissions are involved, test with the smallest useful permission set first.
Configure skills or extension workflows based on the currently available project notes and download page.
Start with verifiable chat, development, file, and model workflows before adding more integrations.
Review the current release status, system requirements, and source notes before installing or waiting for a verified entry.
Check Download Status