←新闻详情

Ollama Blog更新于 10月14日 12:28

Ollama 发布 OpenClaw 最简安装教程

The simplest and fastest way to setup OpenClaw · Ollama Blog

AI 导读

Ollama 官方博客给出用 Ollama 快速安装个人 AI 助手 OpenClaw 的教程,需要 Ollama 0.17 或更高版本、Node.js 以及 Mac 或 Linux 系统,Windows 用户可通过 WSL 安装。运行命令后 Ollama 会自动检测并安装配置 OpenClaw,随后即可在终端中对话;选择 Ollama 云端模型时会自动安装网页搜索插件,本地模型则无需额外插件。OpenClaw 可连接 WhatsApp、Telegram、Slack、Discord 或 iMessage,官方建议智能体使用至少 64k 上下文长度,并提醒在隔离环境中运行以控制文件读取和操作执行的风险。

推荐理由

AI 评分 54

Ollama 官方给出三步安装 OpenClaw 的流程,并说明模型选择与安全隔离要求,适合想本地跑个人 AI 助手的读者参考。

主题

标签

正文

当前显示原文,中文全文译文尚未就绪。

February 23, 2026

OpenClaw is a personal AI assistant that can clear your inbox, send emails, manage your calendar, and complete other tasks via messaging apps like WhatsApp, Telegram, iMessage, or any chat app you already use.

What you’ll need

Ollama 0.17 or later; Node.js (npm is used to install OpenClaw); Mac or Linux system (Windows users can install OpenClaw via WSL - Windows Subsystem for Linux)

Step 1: Run the command

Open a terminal, and type in:

Note: Other models can also be configured. See ollama launch openclaw for recommended models.

Ollama handles everything from here.

Step 2: Install OpenClaw

If OpenClaw isn’t already on your system, Ollama detects that and prompts you for installation.

Ollama will automatically install and configure OpenClaw for you.

Step 3: Start chatting

This is it! OpenClaw will open up in the terminal where you can start chatting with your AI assistant.

Web search

If you selected a model from Ollama’s cloud, Ollama installs the web search plugin automatically. This allows OpenClaw to perform web searches for the latest and up-to-date information.

Local models work out of the box without additional plugins.

Connect messaging apps

OpenClaw can connect to messaging platforms like WhatsApp, Telegram, Slack, Discord, or iMessage to chat with your models from anywhere.

After configuring, make sure you select Finished to save your desired configuration.

Selecting a model

Ollama shows a model selector with recommended and available models. Agents work best with at least 64k context length.

Ollama’s cloud models have full context length, which provide the best experience with agents like OpenClaw.

Cloud models:

Local models (requires GPU VRAM):

Running securely

OpenClaw has the ability to read files and execute actions when tools are enabled. Ensure that you run OpenClaw in an isolated environment and are aware of the risks with giving OpenClaw access to your system.

See the OpenClaw security documentation for more details.

Learn more

Ollama’s cloud models; OpenClaw documentation; Gateway security

来源

Ollama Blog