freeride

You can use free OpenRouter models with OpenClaw by running an automated configuration that discovers models with zero prompt pricing, ranks them by context length and provider trust, and updates ~/.openclaw/openclaw.json as the primary model.

To switch OpenClaw models automatically during rate limits, you configure ranked fallback models in your deployment. This approach rotates to the next available free model to ensure continuous AI responses without manual intervention.

The best way to find free AI models on OpenRouter is to query the API data and filter for models where pricing.prompt equals zero or the model name ends with a :free suffix, then rank them by recency and capabilities.

Configuring free OpenRouter models preserves your existing OpenClaw gateway and environment settings. The setup updates only the primary model, fallbacks, and models allowlist within ~/.openclaw/openclaw.json without overwriting your current configuration.

Free OpenRouter models are ranked by evaluating their context length, supported capabilities, recency of updates, and provider trust score. This ranking ensures the most capable and reliable models are selected as primary and fallback options.

To manage and switch free models in OpenClaw, you use commands like 'auto' to configure the best model and fallbacks, 'list' to view available options, and 'switch' to manually change models before restarting the gateway.