> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# eval • AI Agent SDK

> PraisonAI Agents Evaluation Framework.

# eval

<Badge color="blue">AI Agent</Badge>

PraisonAI Agents Evaluation Framework.

Provides comprehensive evaluation capabilities for AI agents with zero performance
impact when not in use through lazy loading.

Evaluator Types:

* AccuracyEvaluator: Compare output against expected output using LLM-as-judge
* PerformanceEvaluator: Measure runtime and memory usage
* ReliabilityEvaluator: Verify expected tool calls are made
* CriteriaEvaluator: Evaluate against custom criteria

Example:
\>>> from praisonaiagents.eval import AccuracyEvaluator
\>>> evaluator = AccuracyEvaluator(
...     agent=my\_agent,
...     input\_text="What is 2+2?",
...     expected\_output="4"
... )
\>>> result = evaluator.run(print\_summary=True)

## Import

```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
from praisonaiagents import eval
```

### Constants

| Name            | Value                                                                                                                                                                                                         |
| --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `_LAZY_IMPORTS` | `{'BaseEvaluator': ('base', 'BaseEvaluator'), 'AccuracyEvaluator': ('accuracy', 'AccuracyEvaluator'), 'PerformanceEvaluator': ('performance', 'PerformanceEvaluator'), 'ReliabilityEvaluator': ('reliabil...` |

***

## Related Documentation

<CardGroup cols={2}>
  <Card title="Evaluation Concept" icon="gavel" href="/docs/concepts/evaluation" />

  <Card title="LLM as Judge" icon="scale-balanced" href="/docs/features/llm-as-judge" />

  <Card title="Evaluation Loop" icon="rotate" href="/docs/eval/evaluation-loop" />
</CardGroup>