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

# Accuracy

> Ateve delivers leading search accuracy and relevant web context for AI agents.

## Benchmark results

<Tabs>
  <Tab title="SimpleQA">
    <img src="https://mintcdn.com/ateve-f09deb12/k0HIIkRnF5Fo5r8z/images/benchmarks/simpleqa.svg?fit=max&auto=format&n=k0HIIkRnF5Fo5r8z&q=85&s=2e5505550093cf418fd66a7544e43d9e" alt="SimpleQA accuracy: Ateve 95.19%, Exa 94.38%, Brave 87.31%, Tavily 84.40%" width="720" height="340" data-path="images/benchmarks/simpleqa.svg" />
  </Tab>

  <Tab title="FreshQA">
    <img src="https://mintcdn.com/ateve-f09deb12/k0HIIkRnF5Fo5r8z/images/benchmarks/freshqa.svg?fit=max&auto=format&n=k0HIIkRnF5Fo5r8z&q=85&s=a64b262c0662f2e4b38e9d3ff628691b" alt="FreshQA accuracy: Ateve 61.34%, Exa 59.16%, Brave 55.46%, Tavily 56.47%" width="720" height="340" data-path="images/benchmarks/freshqa.svg" />
  </Tab>
</Tabs>

## About this benchmark

This benchmark evaluates how accurately models answer factual questions using retrieved web information, based on SimpleQA and FreshQA. SimpleQA contains short, fact-seeking questions, while FreshQA includes questions about changing facts and false premises.

## Methodology

* Dataset: SimpleQA (4,326 questions) and FreshQA (595 questions: 495 TEST + 100 DEV).
* Answer generation: Answers are grounded in search results retrieved by each provider.
* Scoring: Accuracy (correct answers / total questions), reported separately for each dataset.
* Grading: OpenAI’s official SimpleQA grading prompt and FreshQA’s strict grading rules.
* Retrieval: Up to 10 search results per query. FreshQA results exclude the same five questions with disputed grading for all providers.

## Read more

<Card title="How We Evaluate Search Quality" href="https://ateve.ai/blog/ateve-eval-methodology-search-quality">
  Read the full methodology, evaluation results, and analysis on the Ateve blog.
</Card>