> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tryxlr8.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Changelog

> Product updates and announcements

<Update label="September 17, 2026" tags={["Improved"]}>
  In Query Generation, click a query's source badge to see why it was generated. When the source is an uploaded file, **Source document** links to download that document.

  Importing a categories-and-queries CSV now tags each query automatically. The sample and export include an optional **Citations** column; if you leave it out, tags come from your company briefing, documents, and Keyword Planner. Fill in company details first — or re-import after you do — so those tags can be enriched.
</Update>

<Update label="September 16, 2026" tags={["New","Improved","Fixed","Deprecated"]}>
  Claude Opus 4.1 is no longer available when you generate or optimize content. Use Claude Opus 4.7 or another current model instead.

  Content library rows, filters, and item labels now show the exact model version (for example, Claude Opus 4.7) instead of a generic family name.

  From a generated content piece, you can select comment threads and click **Add to brand learnings** so the feedback can be reviewed and applied to future content. Threads already saved show **Pending**, **In learnings**, or **Dismissed**.

  Comment thread status badges now wrap under the title so they stay fully visible in the sidebar.
</Update>

<Update label="September 15, 2026" tags={["New","Improved","Fixed"]}>
  When an invoice needs payment, [Settings → Billing](/account/billing) now offers **Stripe invoice** (opens Stripe's hosted invoice) alongside **Pay on XLR8** for in-app checkout. Open invoices in Invoice history can show the same Stripe link next to receipts.

  If Google Search Console or Bing Webmaster expires your connection, [Indexing](/indexing/overview) now shows **Reconnect required** and a **Reconnect** button so you can authorize again and restore access.

  After you analyze a domain in [Query Generation](/query-generation), suggested **company aliases** and **tracked competitors** appear on the Categories step for every subtype. They are saved with the query set and copied onto new experiments for mention tracking. For Competitor Alternatives and Competitor Comparison, tracked competitors also match the competitor names on that step.
</Update>

<Update label="September 14, 2026" tags={["Improved"]}>
  New experiments now turn **Enforce Web Search** on by default when you choose GPT Instant or GPT Thinking, so those models search the web before answering unless you switch it off.
</Update>

<Update label="September 14, 2026" tags={["Fixed"]}>
  In Query Generation, queries you add or edit by hand now stay on the list when you regenerate suggestions, instead of disappearing.
</Update>

<Update label="September 11, 2026" tags={["Improved"]}>
  On sentiment **Detailed Analysis**, the **Pros** and **Cons** filter now shows your company name instead of a generic Self label, so you can tell at a glance which phrases are about your brand versus competitors.
</Update>

<Update label="September 11, 2026" tags={["New"]}>
  Keyword Planner in Query Generation now lets you choose a **Language** and one or more **Locations** for search-volume and competition data. Targeting defaults to English and the United States, and is saved with the query set so it is reused when you reopen or copy a briefing.
</Update>

<Update label="September 11, 2026" tags={["New"]}>
  Sentiment **Detailed Analysis** now lets you filter **Pros** and **Cons** with a **Self** / **Competitor** toggle. Self (the default) shows phrases about your brand; Competitor shows phrases attributed to other brands, with brand labels. **Export to CSV** follows the same filter.
</Update>

<Update label="September 10, 2026" tags={["New"]}>
  You can now choose how many queries to generate for each category when you create or regenerate a query set. Set **Queries per category** from 1 to 20 (default 5) on the Categories and Queries steps in Query Generation.
</Update>

<Update label="September 10, 2026" tags={["Improved"]}>
  In **Settings → Tracked URLs**, you can paste a list of URLs (one per line or separated by commas) and add them to multiple visibility or sentiment experiments in one submit. An optional bucket applies to every URL, or XLR8.ai assigns one per URL when you leave it empty.
</Update>

<Update label="September 10, 2026" tags={["New"]}>
  On Analytics **Themes**, you can now filter clustered phrases by aspect, modifier, and intensity for the current Pros or Cons view. Filters apply together at the phrase level, so theme lists and the cluster map only show matching phrases.
</Update>

<Update label="September 10, 2026" tags={["New"]}>
  **Reporting** now combines Google Search Console, Google Analytics, and monthly first-touch survey data in one place. Connect Google, choose your properties, and upload a CSV per month to see search performance, site traffic, and how people say they found you.
</Update>

<Update label="September 10, 2026" tags={["Improved"]}>
  Publishing to Webflow or WordPress now saves as a draft by default, so you can review the CMS item before it goes live. Turn on **Publish immediately** when you are ready, or use **Publish Draft** later — that action defaults to going live. Updating an existing Webflow item also keeps its current draft or live status instead of publishing it automatically.
</Update>

<Update label="September 9, 2026" tags={["Improved"]}>
  Organization members can now approve pending **AI Brand Learnings** and edit their text in Settings. Dismissing a learning, adding one manually, importing from a document, and deleting remain limited to admins.
</Update>

<Update label="September 9, 2026" tags={["Improved"]}>
  Query set rows on Query Generation are fully clickable — click anywhere on a row to open the set. Download CSV and archive appear when you hover a row, and those actions stay on the list.
</Update>

<Update label="September 8, 2026" tags={["New"]}>
  You can now generate and optimize **PDF** content — reports, whitepapers, and other document-style pieces — then download the finished work as a letter-sized PDF from the content editor, alongside Markdown and HTML.
</Update>

<Update label="September 7, 2026" tags={["New"]}>
  The theme cluster map on Analytics now opens in 3D so you can rotate clusters on larger sentiment experiments. Switch to 2D when you want to pan and zoom.

  When you view **Cons**, a **Cons per theme** chart shows how many unique criticisms fall in each cluster. Outliers stay hidden until you turn on **Outliers**.
</Update>