时地方

This commit is contained in:
coco
2026-07-21 14:53:24 +08:00
parent 8b94d76799
commit 1b8027e61c
308 changed files with 61360 additions and 10 deletions
@@ -0,0 +1,159 @@
---
title: "With Razer AIKit, make your agentic AI assistant OpenClaw free. Here's how:"
source: "https://x.com/RazerAI/status/2047284997569368265"
author:
- "[[@RazerAI]]"
published: 2026-04-23
created: 2026-04-24
description: "Run OpenClaw, previously known as Moltbot, locally on the Razer Blade 18 to monitor tasks in real-time and automate workflow actions in the ..."
tags:
- "clippings"
---
![图像](https://pbs.twimg.com/media/HF6TBjEbYAAKPxF?format=jpg&name=large)
Run OpenClaw, previously known as Moltbot, locally on the Razer Blade 18 to monitor tasks in real-time and automate workflow actions in the background. Install and configure OpenClaw, then use Razer AIKit to enable low-latency, on-device AI without cloud dependency. The Razer Blade 18 is built for sustained, high-performance workloads., keeping performance stable while data stays local and setup remains minimal – so the hardware does the work while you build.
[Click here to read this article on Razer AI's website.](https://www.razer.ai/razer-blade-18-openclaw/?utm_source=XTwitter&utm_medium=social&utm_campaign=OpenClawBlade18&utm_content=XArticle)
## 🛠️ Prerequisites for Installing OpenClaw
**Step 1: System Requirements** Ensure WSL on Windows 11 or Ubuntu 24.04 is installed on your Blade 18 along with the latest drivers.
**Step 2: Install Docker** Docker Installation:
```bash
# Add Docker's official GPG key:
sudo apt update
sudo apt install ca-certificates curl
sudo install -m 0755 -d /etc/apt/keyrings
sudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc
sudo chmod a+r /etc/apt/keyrings/docker.asc
# Add the repository to Apt sources:
sudo tee /etc/apt/sources.list.d/docker.sources <<EOF
Types: deb
URIs: https://download.docker.com/linux/ubuntu
Suites: $(. /etc/os-release && echo "${UBUNTU_CODENAME:-$VERSION_CODENAME}")
Components: stable
Signed-By: /etc/apt/keyrings/docker.asc
EOF
sudo apt update
sudo apt install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin
sudo usermod -aG docker $USER
newgrp docker
```
**Step 3: Install NVIDIA Container Toolkit**
```bash
nvidia-ctk runtime configure --runtime=docker --config=$HOME/.config/docker/daemon.json
systemctl --user restart docker
sudo nvidia-ctk config --set nvidia-container-cli.no-cgroups --in-place
```
**Step 4: Verify Installation**
```bash
docker -v
```
![图像](https://pbs.twimg.com/media/HGkuVAZbIAAG5Dc?format=jpg&name=large)
## 🧠 Razer AIKit Installation & Inferencing
**Step 1: Pull & Start AIKit Container**
```bash
mkdir -p $HOME/.cache/huggingface
docker run -it \
--restart=unless-stopped \
--gpus all \
--ipc host \
--network host \
--mount type=bind,source=$HOME/.cache/huggingface,target=/var/aikit/.cache/huggingface \
razerofficial/aikit:latest
```
**Step 2: Download Model** Once inside the container, run:
```bash
rzr-aikit model download openai/gpt-oss-20b
```
> Note: This guide uses openai/gpt-oss-20b. You can also try other models like [DeepSeek-R1-Distill-Qwen-7B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B) or [Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B)**,** available on [Hugging Face.](https://huggingface.co/)
> If your selected model exceeds the memory capacity of a single Razer Blade 18, use [Razer AIKit](https://github.com/razerofficial/aikit), our open-source solution, to optimize the load across a cluster of GPUs. Further scale your local AI workstation with our [Razer eGPU](https://www.razer.com/sg-en/gaming-egpus/razer-core-x-v2?srsltid=AfmBOoqqoGiX_biLercjwgcd-PyuD3UUuHPaPLEt-JES2U73-H9WaTVW) and [Forge AI Dev Workstation](https://www.razer.ai/forge-workstation/?utm_source=XTwitter&utm_medium=social&utm_campaign=OpenClawBlade18&utm_content=XArticle) to increase GPU resources for higher-level inferencing.
**Step 3: Start the Model Server**
```bash
rzr-aikit model run openai/gpt-oss-20b
```
> **Note**: If running on WSL, include the flag --gpu-memory-utilization 0.8
**Step 4: Test Model Inference (Optional)** To confirm the model is running correctly, execute:
```bash
rzr-aikit model generate "Tell me about Telegram Bot - what is it?"
```
![图像](https://pbs.twimg.com/media/HGkuggpa0AA-xD6?format=jpg&name=large)
## 🦞 OpenClaw Installation & Configuration
**Step 1: Install OpenClaw**
```bash
curl -fsSL https://openclaw.ai/install.sh | bash
```
**Step 2: Follow Setup Instructions to Connect the Local Model:**
1. Select \[Quickstart\] -> \[Custom Provider\]
2. Enter the OpenAPI endpoint for your local model (default: http://127.0.0.1:8000/v1)
3. Choose \[Paste API key now\] and leave it blank.
![图像](https://pbs.twimg.com/media/HGkuchda4AAs0mb?format=jpg&name=large)
1. Choose \[OpenAPI-compatible\] and enter model ID (openai/gpt-oss-20b)
2. Select Telegram channel, and proceed to Telegram bot setup
![图像](https://pbs.twimg.com/media/HGkumIla0AEVucV?format=jpg&name=large)
## 📱 Telegram Bot Setup
**Step 1: Create a Telegram Bot Token**
1. Open Telegram and search for [@BotFather](https://x.com/@BotFather).
2. Run the following command: /newbot
3. Copy the token you receive, it will look like: 123456789:AAExampleTokenHere
**Step 2: Add Telegram Token to OpenClaw**
1. Select \[Enter Telegram Bot Token\] and paste your token
2. Finish the rest of the openclaw setup flow based on your preferences
**Step 3: Text Your Bot on Telegram!**
1. Open Telegram
2. Search for your new bot’s username
3. Send a message
4. Bot will request to authenticate; run the command below:
```bash
openclaw pairing approve telegram <code>
```
Your Personal AI Assistant OpenClaw is good to go!
<video preload="none" tabindex="-1" playsinline="" aria-label="嵌入式视频" poster="https://pbs.twimg.com/amplify_video_thumb/2047209395344564224/img/0-VZoFkungf3gljC.jpg" style="width: 100%; height: 100%; position: absolute; background-color: black; top: 0%; left: 0%; transform: rotate(0deg) scale(1.005);"><source type="video/mp4" src="blob:https://x.com/9cefa826-2e3c-44bf-b984-e7e43365da2f"></video>
![](https://pbs.twimg.com/amplify_video_thumb/2047209395344564224/img/0-VZoFkungf3gljC.jpg?name=large)
Explore 50+ other Automations enabled by OpenClaw [here](https://openclaw.ai/integrations)
> **Note:** OpenClaw is an experimental tool and not directly affiliated with Razer. Please ensure your environment is properly secured. Razer is not liable for data loss or issues arising from its use.