> For the complete documentation index, see [llms.txt](https://roborus.gitbook.io/whitepaper/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://roborus.gitbook.io/whitepaper/features-and-capabilities/6.7-rewards-for-running-ai-models-locally.md).

# 6.7 Rewards for Running AI Models Locally

Most robotics platforms rely solely on robots for task execution, leaving non-robot owners with little opportunity to participate. Roborus breaks this limitation by enabling **any user with a computer** to contribute to the network by running AI models locally.

Here’s how it works:

* Users install the **Roborus Agent SDK** on their systems.
* The SDK downloads lightweight AI models or simulations for testing.
* Their system resources (CPU, GPU, network) are used to run and validate models.
* If the test passes, the user earns **token rewards**.

This approach serves multiple purposes:

1. **Decentralized compute network**: Roborus leverages thousands of distributed machines instead of relying on centralized servers.
2. **Scalable model testing**: New robotics and AI models can be tested at scale across diverse environments.
3. **Democratized earning**: Anyone can participate, even without owning a robot.

The impact is a **broader ecosystem** where everyday users can earn rewards and contribute value, creating a more inclusive platform.