> For the complete documentation index, see [llms.txt](https://docs.geopard.tech/geopard-tutorials/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.geopard.tech/geopard-tutorials/product-tour-web-app/geopard-ai-assistant/accuracy-and-evaluation.md).

# Accuracy and evaluation

We measure the assistant's agronomic accuracy on a versioned benchmark. We publish every category score, including areas still improving.

### The benchmark

* 503 expert-level agronomy questions, CCA-style.
* Five categories: soil and water, crop management, nutrient management, pest management and IPM, precision ag specialty.
* Versioned and re-run on releases. Current version: v2, July 2026.

### Current scores

| Category                | Score |
| ----------------------- | ----- |
| Precision ag specialty  | 100%  |
| Soil and water          | 95.8% |
| Crop management         | 93.8% |
| Nutrient management     | 92.9% |
| Pest management and IPM | 77.5% |

### Why the question set is private

Public benchmark questions leak into AI training data. Once that happens, scores stop measuring capability. The question set stays private so future scores remain meaningful. We publish the methodology, all category scores, and sample questions on request.

### Validation beyond the benchmark

Benchmark scores measure knowledge. Every generated plan is dry-run validated before creation. Checks cover nutrient balance and removal coefficients against extension recommendations. They cover soil-test critical levels and build-up conventions from your lab. They also cover application rate limits, setback distances, and knowledge base rules. If a check fails, no map is generated.

### Your role

The assistant supports professional judgment. It does not replace a certified local agronomist. You review and approve every prescription. Validate recommendations against your crop plan, soil tests, product labels, and local regulations.

### Related

* [GeoPard AI Assistant overview](/geopard-tutorials/product-tour-web-app/geopard-ai-assistant.md)
* [AI Assistant FAQ](/geopard-tutorials/product-tour-web-app/geopard-ai-assistant/precision-agriculture-ai-assistant-faq.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.geopard.tech/geopard-tutorials/product-tour-web-app/geopard-ai-assistant/accuracy-and-evaluation.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.