> ## 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.

# Latency

> Built for real-time experiences, with sub-second retrieval optimized for production workloads.

## Benchmark results

<img src="https://mintcdn.com/ateve-f09deb12/k0HIIkRnF5Fo5r8z/images/benchmarks/latency.svg?fit=max&auto=format&n=k0HIIkRnF5Fo5r8z&q=85&s=6dc64d143ee209edb5434d8d68eb937a" alt="Homepage latency comparison showing the p50 to p95 range for Ateve, Exa, Tavily, and Brave on a 0–4000 millisecond axis" width="720" height="590" data-path="images/benchmarks/latency.svg" />

## About this benchmark

This benchmark compares accuracy and response time (latency) for different providers on the SimpleQA dataset, demonstrating the trade-offs between speed and answer correctness.

## Methodology

* Latency: Measured as median (p50), including p95 range for tail latency
* Accuracy measured on SimpleQA

## Read more

<Card title="How We Measure Search Speed" href="https://ateve.ai/blog/ateve-latency-eval-methodology-search-speed">
  Read the full methodology, measurement definitions, and results on the Ateve blog.
</Card>
