---
title: "Inference Service model performance metrics"
sidebar_label: "Model performance metrics"
sidebar_position: 3
description: "Description of Inference Service model performance metrics"
---

import { CustomTable } from '@selectel/docux/components';
import Formbricks from '@theme/MDXComponents/Formbricks';

# Inference Service model performance metrics

Using metrics, you can compare model performance across different configurations and select the one you need. Metrics are displayed in the card for each [configuration](/foundation-models-catalog/create/configurations.mdx) when [creating an inference service](/foundation-models-catalog/create/create-inference-service.mdx).

<CustomTable>
  <table>
    <tbody>
      <tr>
        <th>Avg Time to First Token</th>

        <td>
          Average time from request receipt to generation of the first token in milliseconds
        </td>
      </tr>

      <tr>
        <th>Avg Request Throughput</th>

        <td>
          Average number of requests processed per second
        </td>
      </tr>

      <tr>
        <th>Output Token Throughput</th>

        <td>
          Average number of generated tokens per second
        </td>
      </tr>

      <tr>
        <th>Request Latency</th>

        <td>
          Average time from request receipt to full response completion in seconds
        </td>
      </tr>
    </tbody>
  </table>
</CustomTable>

<Formbricks />