<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title><![CDATA[Litmiq Blog]]></title><description><![CDATA[Litmiq Blog]]></description><link>https://litmiq.hashnode.dev</link><generator>RSS for Node</generator><lastBuildDate>Sat, 21 Nov 2026 18:40:56 GMT</lastBuildDate><atom:link href="https://litmiq.hashnode.dev/rss.xml" rel="self" type="application/rss+xml"/><language><![CDATA[en]]></language><ttl>60</ttl><item><title><![CDATA[OpenAI Evals shuts down on 30 November 2026: export first, then migrate]]></title><description><![CDATA[Originally published on Litmiq. Updated 13 November 2026. Sources on that page were read the same day. OpenAI Evals became read-only on 31 October 2026 and…]]></description><link>https://litmiq.hashnode.dev/openai-evals-shuts-down-on-30-november-2026-export-first-then-migrate</link><guid isPermaLink="true">https://litmiq.hashnode.dev/openai-evals-shuts-down-on-30-november-2026-export-first-then-migrate</guid><category><![CDATA[AI]]></category><category><![CDATA[llm]]></category><category><![CDATA[openai]]></category><category><![CDATA[Testing]]></category><dc:creator><![CDATA[Litmiq]]></dc:creator><pubDate>Tue, 17 Nov 2026 22:19:06 GMT</pubDate><content:encoded><![CDATA[<p><em>Originally published on <a href="https://litmiq.com/migrate-from-openai-evals">Litmiq</a>. Updated 13 November 2026. Sources on that page were read the same day.</em></p>
<p>OpenAI Evals became <strong>read-only on 31 October 2026</strong> and is scheduled to <strong>shut down on 30 November 2026</strong> (dashboard, API, and graders for eval workflows). OpenAI recommends migrating to Promptfoo. Dates and wording should be checked against <a href="https://developers.openai.com/api/docs/deprecations">OpenAI's deprecations page</a>, which is authoritative if anything here disagrees.</p>
<p>This is a short export-first path. The full guide, including the export script and an in-browser converter, is at <a href="https://litmiq.com/migrate-from-openai-evals">litmiq.com/migrate-from-openai-evals</a>.</p>
<h2>The short answer</h2>
<ul>
<li><strong>Export first.</strong> There is no export button, but the Evals API still returns your evals, runs, and output items while the product is read-only. Save all of them, read-only, before 30 November 2026.</li>
<li><strong>Convert.</strong> The Litmiq guide includes a free, in-browser converter that turns that export into a <code>promptfooconfig.yaml</code>, a CSV, or a Litmiq suite.</li>
<li><strong>Re-check scores.</strong> Similarity and model-graded scores will not match OpenAI's exactly. Re-run a sample and re-tune thresholds before you trust them.</li>
</ul>
<p>A close, separate deadline from the same deprecations page (read 13 November 2026): <strong>gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, o3 and o3-pro</strong> are listed as shutting down on <strong>11 December 2026</strong>. If an eval pinned one of those, re-baseline on the replacement OpenAI names.</p>
<h2>Timeline (as published by OpenAI, read 13 Nov 2026)</h2>
<table>
<thead>
<tr>
<th>Date</th>
<th>What happens</th>
</tr>
</thead>
<tbody><tr>
<td>3 June 2026</td>
<td>OpenAI publishes the deprecation notice for the Evals dashboard, API and graders for eval workflows, and for <code>/v1/prompts</code> reusable prompt objects.</td>
</tr>
<tr>
<td>31 October 2026</td>
<td>Evals become read-only for existing users.</td>
</tr>
<tr>
<td>30 November 2026</td>
<td>The Evals platform is scheduled to shut down. OpenAI recommends Promptfoo. Reusable prompt objects are deprecated the same day — move prompt text into your app code.</td>
</tr>
<tr>
<td>11 December 2026</td>
<td>Separate but close: several GPT-5 / o3 models shut down. Re-baseline if you pinned them.</td>
</tr>
</tbody></table>
<p>If OpenAI changes these dates, their pages win.</p>
<h2>Where to go after you export</h2>
<ul>
<li><strong>Promptfoo</strong> — this is OpenAI's recommended replacement, and the right default for CLI and CI workflows.</li>
<li><strong>A browser UI</strong> — if you want to inspect cases side by side without installing anything, you can load a converted suite in <a href="https://litmiq.com/app">Litmiq</a>. Litmiq is free, you bring your own OpenAI key, and it does not replace Promptfoo as a merge gate. I built Litmiq; the longer guide is explicit about those limits.</li>
</ul>
<p>DeepEval and other code-first tools remain reasonable if that is already your stack. The point is not to switch tools twice: export once, then pick the workflow you will actually run.</p>
<h2>What will not come across cleanly</h2>
<p>Model-graded and similarity scores from OpenAI's graders are not something you should copy-paste as pass/fail thresholds in a new tool and trust. Re-run a sample. Treat old scores as history, not as a grade book that still applies.</p>
<p>The in-browser converter and the export steps (including a worked IT-ticket example) are on the <a href="https://litmiq.com/migrate-from-openai-evals">canonical guide</a>. I am not pasting a GitHub repo here, because the public converter repository is not published yet; use the guide on the site.</p>
<h2>A short checklist</h2>
<ol>
<li>Export every eval, run, and output item you still care about, while the API is read-only and still answering.</li>
<li>Keep the raw export, even after you convert it.</li>
<li>Convert to Promptfoo YAML and/or a CSV / Litmiq suite.</li>
<li>Re-run a sample on the models you will actually use after 11 December 2026.</li>
<li>Re-tune thresholds. Do not ship a CI gate on un-checked migrated scores.</li>
<li>Move any reusable prompt objects out of <code>/v1/prompts</code> and into your own source.</li>
</ol>
<p>If you find an error in the dates or the export steps, the <a href="https://litmiq.com/migrate-from-openai-evals">canonical page</a> is the best place to flag it — that is the document I will correct first.</p>
]]></content:encoded></item></channel></rss>