freeride

What This Skill Does

FreeRide is a specialized utility for the OpenClaw ecosystem designed to optimize AI model selection while minimizing operational costs. By leveraging the OpenRouter API, FreeRide automates the complex task of selecting, ranking, and configuring high-quality, free-tier AI models. It acts as an intelligent layer that sits between your OpenClaw configuration and the available model landscape, ensuring that your agent always has a performant "primary" model, supported by a chain of robust fallbacks.

When you invoke FreeRide, it performs a systematic analysis of available free models based on context window length, inherent capabilities, model recency, and provider reliability. It then safely modifies your ~/.openclaw/openclaw.json file to inject these preferences without disturbing your existing gateway or plugin configurations. This automated approach ensures that if a specific model encounters rate limits or downtime, OpenClaw transparently transitions to the next best candidate in your configured fallback sequence.

Installation

To integrate FreeRide into your workflow, ensure you have the OpenClaw CLI installed, then execute the following command in your terminal:

npx clawhub@latest install freeride

Alternatively, you can install it from the source repository via:

clawhub install openclaw/skills/skills/shaivpidadi/freeride

Once installed, remember to set your OpenRouter API key in your environment variables using export OPENROUTER_API_KEY="sk-or-v1-...".

Use Cases

FreeRide is perfect for users looking to maximize the utility of their AI agent without incurring subscription costs. It is highly recommended for developers who frequently hit rate limits, users experimenting with diverse model capabilities, and those who want to automate their environment management to spend less time configuring tools and more time building. Whether you are scaling an application or performing deep research, FreeRide keeps your agent functional through intelligent model switching.

Example Prompts

1. "OpenClaw, I'm hitting rate limits constantly. Can you set up FreeRide to handle my fallbacks automatically?"
2. "Switch my primary model to Qwen3-coder and make sure I have at least 5 fallbacks configured for when it goes down."
3. "I want to reduce my AI costs. Use the freeride skill to reconfigure my openclaw.json with the best available free models."

Tips & Limitations

• Proactive Management: Always use freeride status periodically to ensure your fallback chain is still valid.
• Refresh Often: Use freeride refresh occasionally to fetch the latest ranking data, as the AI ecosystem changes rapidly.
• Safety Note: FreeRide performs direct modifications to your local JSON configuration files. Always ensure your environment is backed up before performing massive reconfigurations.
• Rate Limits: While fallbacks mitigate downtime, they do not bypass OpenRouter's global rate limits. If you exhaust your limits for all free tiers, you may still need to wait before subsequent requests.