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

# Playground timeout with long-running model requests

> [!NOTE]
>
> **Applies to:**
>
> * Plan - Any
> * Deployment - Braintrust-hosted
> *
> * Use case - Playground runs with long-running models (85+ seconds execution time)

## Summary

**Issue:** Playground runs with long-running models like `gpt-5-mini` (85+ seconds) appear to hang or timeout, showing no progress updates in the UI even though the request completes successfully in the background.

**Cause:** Server-Side Events (SSE) connection drops before the job completes, preventing the UI from receiving completion updates.

**Resolution:** Refresh the page to view completed results

## Resolution Steps

### Workaround for current issue

#### Step 1: Wait for expected completion time

Allow sufficient time for the model to complete processing (typically 85+ seconds for `gpt-5-mini`).

#### Step 2: Refresh the browser page

Reload the page to display the completed results that processed in the background.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.