A powerful CLI-driven skill providing unified access to over 100 leading large language models through OpenRouter integration.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install llm-models
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install llm-models using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is LLM Models via OpenRouter?
The LLM Models skill via OpenRouter offers developers and automation enthusiasts a streamlined interface to the world's most capable AI models. By utilizing the inference.sh CLI, this Openclaw Skills integration allows you to switch between frontier models like Claude 4.5, Gemini 3 Pro, and Kimi K2 with a single command.
This skill eliminates the friction of managing multiple API keys or navigating complex SDKs for different providers. It focuses on versatility and efficiency, providing automatic fallbacks and cost-optimization features that make it an essential component for building robust AI agents and automated development workflows.
LLM Models via OpenRouter Use Cases
- Automated code generation, debugging, and peer reviews using high-reasoning models like Claude Opus.
- Developing sophisticated AI agents that require multi-step reasoning via Kimi K2 Thinking.
- Cost-effective content generation and translation using lightweight models like Claude Haiku.
- Building research workflows by integrating LLM outputs with search and data analysis tools.
- Creating conversational interfaces and chatbots with access to the latest frontier models.
How LLM Models via OpenRouter Works
- Install the inference.sh CLI to establish the environment for Openclaw Skills.
- Authenticate your session using the login command to connect to the model provider.
- Select a specific model App ID or use the any-model selector for automatic optimization.
- Pass a JSON-formatted input containing your prompt and optional system instructions.
- Receive the model's response directly in your terminal or pipe it into subsequent automation steps.
LLM Models via OpenRouter Setup
To get started with this skill, install the CLI and authenticate your account:
curl -fsSL https://cli.inference.sh | sh && infsh login
Once logged in, you can run any supported model immediately. For example, to use Claude Sonnet for a reasoning task:
infsh app run openrouter/claude-sonnet-45 --input '{"prompt": "Explain quantum computing"}'
LLM Models via OpenRouter Data Schema & Taxonomy
The skill processes data using a structured JSON input schema to ensure consistent interactions across different Openclaw Skills models.
| Property | Type | Description |
|---|---|---|
prompt |
String | The main instruction or question for the LLM. |
system |
String | Optional context to define the AI's behavior or persona. |
input.json |
File | Local file containing prompt parameters for complex queries. |
All responses are returned as standard output, which can be saved to local files or processed by other CLI utilities.
LLM Models via OpenRouter Advanced Features
- Auto-selection of the most cost-effective model using the any-model App ID.
- Seamless integration with other Openclaw Skills for RAG (Retrieval-Augmented Generation) workflows.
- Programmatic control via the Agent SDK for building autonomous multi-agent systems.
- Input sampling to quickly generate and save JSON templates for complex prompt engineering.
- High-performance inference with automatic fallback to ensure reliability in production environments.
SKILL.md
Loading
Related Openclaw Skills
Landing Page Design
A framework for building high-converting landing pages using AI-driven asset generation and proven marketing psychology.
inference-sh
A unified command-line interface for executing 150+ serverless AI applications across image, video, text, and social media domains.
Image Upscaling
A comprehensive AI toolkit for upscaling, restoring, and increasing image resolution via high-performance inference models.
Image to Video Conversion
A high-performance guide and toolset for converting static images into dynamic, cinematic videos using state-of-the-art AI motion models.
Logo Design Guide
A comprehensive framework for generating professional, scalable logo concepts and brand marks using advanced AI image models.
Nano Banana 2 - Gemini 3.1 Flash Image Preview
A high-performance image generation and editing tool leveraging Google Gemini 3.1 Flash Image Preview through a streamlined CLI interface.
METADATA
Featured*
