What's new in LLM Pulse

When more than one AI model reports searches, pick the models you want in Query Fan Out to see only their searches. The filter applies to the By Query and By Prompt views and to the export.

In Citations, open a page's details from the By Page tab to see its citations by AI model and over time, counted with the filters you set on the table.

Query Fan Out now shows the web searches Claude runs while answering your prompts, next to the searches from other models.

Change your login email from Account Settings. Confirm with your password or Google, and with your two-factor code if it is on. The new address takes effect once you open the confirmation link sent to it.

Add up to 10 more domains your company owns, such as a product site or an acquired brand, so citations to them count as your own. Each domain includes its subdomains, and past citations are updated automatically.

Create a GEO audit for a domain or page and run it once, weekly, or monthly. Each run is saved, so you can follow the score over time and see which issues are new, persisting, regressed, or fixed. Every audit and report now sits in one GEO Optimization list you can filter by type, status, and schedule.

The Per URL tab in Agent Analytics is now available on the Scale tier or higher, for every connected log source. It uses a monthly credit allowance shared by your projects, and each project chooses which bots to analyze per URL.

Connect a Matomo site to AI Traffic to see the visitors, visits, conversions, and entry pages that AI assistants send to your website. Enter your Matomo URL, site ID, and an auth token, and the connection is checked before the first sync.

In Reddit Intelligence, click a thread to read its saved text, the brands it mentions, and the AI answers that cite it. Click an author to see the brands in their posts and their cited threads.

In Plans & Billing, buy any number of extra prompts from 10 to 9,900 in one purchase, with shortcuts for 25, 50, and 100. You can also choose how many extra prompts to remove at renewal.

Switch Tag and Country Leadership from rankings to stacked share bars. Compare your brand with the strongest competitors, see where your brand leads, and keep the same view in shared links and PDFs.

Skip links, visible focus states, and keyboard-friendly dialogs and dropdowns make common pages easier to navigate. Charts include screen reader summaries and readable data tables, while table editors, form hints, and errors stay connected to their fields.

Download chart data from each chart, or export the AI Model Insights brand-by-model table through Exports as CSV or Excel. The file follows the selected metric, brands, models, dates, and filters.

When you add or open a long prompt, LLM Pulse names the tracked models that will skip it and shows their character limits. The prompt still runs on the models that support its length.

Execution usage on the Dashboard and Plans and Billing now reflects the current daily, weekly, or monthly tracking cycle. An earlier cycle's usage no longer remains in the count after the next cycle begins.

Copy a page URL to share the current project, date range, filters, search, sort, pagination, visible columns, and chart selections. Teammates with access open the same view.

Prompt creation, CSV imports, bulk tagging, URL groups, AI tags, Run All Reports, and bulk deletion now show what succeeded, what was skipped, and why. Result summaries separate duplicates, invalid rows, missing records, and request failures.

Agent Readiness now detects JSON-LD added after a page loads and lists site-wide and path-specific Content-Signal rules. Relevant checks also show the resolved homepage URL and response status.

Naver AI Briefing and Baidu AI are now add-ons you can activate from Plans and Billing. Both collect on your account's tracking schedule and appear in model filters, comparisons, and charts next to your other AI models.

Move to a higher plan while you are trialing, straight from Change plan. Choose monthly or yearly billing, review the amount due today, and confirm to start the paid plan right away. Your tracking cadence stays the same.

The schema report now describes the markup it actually found on the page and separates required signals from optional ones. A subtype such as Product satisfies the checks for the types it inherits from, so valid markup is no longer reported as missing.

Reputation reports and studies now show which models have already answered and let you read the results so far instead of waiting for every model to finish. You can export the partial data too.

Connecting a data source now happens on one page. Pick your provider from the source list and fill in the connection form in place, without moving between screens. The page also lists the integrations we are working on and lets you tell us which one you need.

Custom Reports has a new Send now action that emails a report to its saved recipients straight away, without changing the scheduled send. Each report in the list shows its recipients, last delivery, date range, and widget count, with view, edit, export, and send actions in one menu.

Brand Reputation reports have a new Moodboard tab that generates six images from your three highest and three lowest scoring attributes. Generate a board from any finished report, then regenerate or hide individual images and bring hidden ones back. The board is included in the report PDF.

Marketing email choices stay in sync across account settings, newsletter links, and campaign emails. You can turn export-ready messages off separately, while one-click unsubscribe works for project digests in supported email clients.

If you refresh project setup or the Add Prompts dialog, the draft returns in the same browser tab. An invalid social link no longer stops the rest of the project from being saved.

The new Brand discovery view suggests companies that compete with your core business. Review suggestions saved during setup or refresh them from the Competitors page.

Change the title or individual text blocks of a completed GEO Writer task without regenerating it. Your edits appear in exports, and you can restore the original AI version.

The Agent now summarizes older exchanges as a conversation grows while keeping recent messages in full. A context indicator appears when space runs low.

ChatGPT Shopping now labels product and merchant prices with their currencies. Price averages and trends use one currency at a time, and the currency filter carries through to exports.

Overview, Citations, Responses, Sources, and the project dashboard open faster, especially on large accounts. Answer previews and Agent history load when opened, and completed background work refreshes without a full page reload.

Connect the Amazon S3 bucket behind your CloudFront access logs. Agent Analytics imports new W3C log files each day, including gzip-compressed files, and still lets you sync on demand.

LLM Pulse now warns when a Google Analytics or Search Console property appears to belong to another project. Switching properties clears the old property's data before the new sync begins.

Choose whether each competitor's citations match every subdomain, one exact host, or one host and path. Changes reprocess past citations so reports follow the new rule.

Search Help Center article titles and summaries from Quick Search in English or Spanish. Product pages still appear first.

See when a plan or model purchase is waiting for 3D Secure approval, then reopen the secure payment step from Billing or email before it expires.

Open the same Brand Book from Project Settings or GEO Writer, then review optional AI suggestions for your industry, audience, voice, products, goals, and other brand context before saving.

When you add a model to your account, LLM Pulse runs your current prompts on it immediately instead of waiting for the next scheduled run.

Update the owned domain from Project Settings, reprocess historical citations for the new website, and optionally keep the old domain as a competitor.

Switch to a compact icon rail and keep that view as you move between pages or reopen the browser.

Use the new YouTube Intelligence report to compare cited channels and videos, measure the share earned by your connected channels, filter the results, and export them.

Open long date ranges in Responses and Citations faster, including workspaces with several projects and large histories.

Find cited pages that speak negatively about your brand, then track negative citations, distinct negative URLs, and newly negative URLs over time.

Build reports for the previous calendar month and add Sentiment by Brand to the report page, public link, scheduled email, PDF, and Excel workbook.

Start with a topic, keyword, query, prompt, or domain and get focused suggestions built around that exact input for the selected market.

See who wrote each message, rate assistant replies, and get more reliable answers about dates, markets, sources, and account data.

Navigate headers, sidebars, tabs, filters, and wide tables on smaller screens without controls running off the page.

Open the latest stored content for a cited page, including YouTube video transcripts, see its crawl date and status, and jump to the live URL when needed.

Hide disabled dimensions from every report view, then create one fresh report per month to apply dimension and competitor changes right away.

Use two new browser tools to clean copied formatting or detect and selectively remove hidden Unicode marks without uploading your text. Find both, plus every free LLM Pulse checker, in the new Free AI Search Tools hub.

Repeated versions of the same creative are now merged, and long ad lists keep their spacing as more results load.

Add Performance KPIs, Citation Rate trends, model breakdowns, Tag Leadership, and Country Leadership to reports, public links, PDFs, and Excel workbooks.

Pair compact half-width sections or use full-width sections across the report page, public link, and PDF, with aligned chart heights.

Open Agent Readiness and Shopping Feed results from GEO Optimization without landing on the wrong page.

Start weekly charts in the correct period, keep useful donut breakdowns, and remove unused position reporting.

See project countries, languages, and competitors in a dedicated information row with clearer KPI definitions.

Control prompt deletion independently from prompt creation, with clear bulk action errors.

Analyze the signals AI agents look for and save the result as an Agent Readiness report.

Find completed files first and launch the available PDF report types from the export page.

Change plans, add models, review itemized pricing before confirming, see renewal totals, and cancel a scheduled change from one place.

Let team members open Recommendations when they can read the report but cannot export it.

Attach an Excel version of report data to scheduled Custom Report emails.

Create Sentiment PDF reports through the shared Export History flow.

Read PDF KPI cards that use the same mention and citation totals as Project Overview.

Read KPI widgets that use the same totals as the study dashboard.

Manage project notification choices through the account email preference center.

Set up time-based and split tests more easily and read comparisons across reports with clearer tables.

Download advertiser, response, merchant, price, product, and shopping response tables.

Focus reports on own-brand, competitor, or non-brand prompts across private, public, PDF, and scheduled views.

Create annotations when prompts, enabled models, or tracking frequency change, with controls to disable or hide them.

Open Alexa for Shopping responses and review product results as structured, linked items instead of flattened raw fields.

Team members with export access can download the Master XLSX, with each sheet limited to the product areas they can view.

Create an account with one password field, built-in requirements, and a control to show or hide what you typed.

Give Sentiment custom classification instructions and rerun analysis with the new guidance.

Repair malformed report output that previously hid competitor brands or analysis sections.

Download report and data exports for a full week instead of 48 hours.

Receive generated sentiment comments using the project's selected language.

Include cited pages that criticize the brand in negative sentiment email summaries.

Download AI Overview, leadership, model comparison, prompt, and execution tables as CSV or Excel.

Download the scores and explanations available in a Custom AI Study.

Protect product counts, prices, and merchant data by keeping distinct products in separate rows.

Add Claude tracking to the same execution and comparison reports used by other models.

See when an AI model answered from its own knowledge instead of using web search.

Account owners see the add-on models they can add and go straight to Plans & Billing to turn them on.

See sample prompts, eligible-country coverage, and more accurate merchant and advertiser ownership.

Preserve inline citation markers while avoiding a duplicated source list in response details.

Configure analyst reports beyond brand reputation through a guided setup and dedicated Studies workspace.

Import prompts with countries, languages, and tags through a previewed migration flow and downloadable template.

Start a new analysis directly from any saved GEO tool report.

Resend a pending invitation from the team list with a controlled resend limit.

Narrow the Integrations page by category and find the Webhooks connection more quickly.

Add a new tag inline while creating prompts.

Paid accounts can add Alexa for Shopping from Plans and Billing. It follows your tracking cadence and appears in prompt tracking, filters, charts, and AI Model Insights.

Read charts through a shared role-based palette and optionally emphasize the project brand.

Keep the Brand Type filter on non-brand prompts by default and carry the saved project choice between reports.

Use one ownership rule for citations, background references, aliases, and sentiment coverage across reports.

Get recommendations shaped by brand context, owner guidance, and the data available in the project.

Manage email categories and unsubscribes from one self-service page in Account Settings.

Save a named group of competitors and apply it to report filters in one action.

Remove selected competitors from both Reputation analysis and the finished report.

Drill from YouTube channel and video rows into related prompts, with a consistent chart palette across Owned Media.

Create PDFs through the same queued history, status notifications, download flow, and expiry model as CSV and Excel.

Compare current and previous periods across AI Model Insights, ChatGPT Ads, Shopping, Owned Media, and Reddit.

See labels and values in the Overview matrix and distribution adapt to the selected metric.

Keep the project brand easy to find in Tag and Country Leadership tables.

Review every available recommendation type together from the default All view.

Purchase capacity add-ons in the app with clear usage and per-unit pricing.

Add custom guidance to a collection so AI-assisted categorization follows the project's needs.

Explore AI Traffic through Overview, Channels, and Landing Pages, with conversions available as a channel metric.

Star the prompts you use most and keep them at the top of the Prompts view.

Compare average mention position across tracked models in AI Model Insights.

Use permission templates or set feature and action access for each teammate, including prompts and invitations.

Send invoice copies to extra email addresses from Billing settings.

Receive one completion email for the all-in-one GEO analysis instead of a separate email for every check.

Explore composition and bandwidth charts, company filters, broader connection options, and per-URL crawl data.

Create prompts with collections, own-brand controls, competitor assistance, and consistent research methods.

Send product events through outbound webhooks or connect LLM Pulse with Zapier, Make, and n8n.

Filter Responses and Citations by whether your brand or selected competitors were mentioned or cited.

Download the YouTube tab as CSV or Excel with support for larger result sets.

Drag columns into the order that best fits your workflow and keep the choice for future visits.

Apply country and language selections across Sentiment views.

Run recommendations with multiple filters and include social sources from TikTok, Facebook, and Instagram.

Focus headline visibility and Share of Voice on generic prompts so brand-name queries do not inflate the result.

Open the LLM Pulse Agent on its own page and copy or export its tables to CSV and Excel.

Use Adobe Analytics or Adobe Customer Journey Analytics as an AI Traffic source.

Compare brand leadership by country alongside the existing tag-based view.

Generate dedicated Sentiment and Reputation strategies with fewer duplicate suggestions.

Get broader analysis across content readiness, crawlability, discoverability, schema, and site structure. Each report now shows the final URL and HTTP status it analysed, and flags partial content or error pages instead of presenting them as complete results.

Start new projects with their detected country instead of a fixed default.

Use an account token with guided account and zone selection in Agent Analytics.

Read model comparison percentages that use the same totals as Project Overview.

Add Piano Analytics as a traffic source and compare it with other analytics connections.

Save brand context in Project Settings so Content Intelligence can produce more useful drafts.

Open completed files directly while duplicate and overlapping export notices stay out of the way.

Create a Master XLSX alongside focused Quick Exports from one place.

Keep the same Sentiment and Citation filters in downloaded data that you used on screen.

Find profile, currency, password, two-factor security, and data connections through clear settings tabs.

Restrict a teammate to selected projects with clear feedback if the change cannot be saved.

Paste prompt lists during onboarding or start from a dedicated Suggested prompts tab.

Paste up to 1,000 domains, hosts, or URLs to narrow Citation results at once.

Explore products, merchants, price insights, trends, and responses in dedicated views.

Use the tracked brand name in charts, exports, reports, and prompt execution while keeping a distinct project label.

Match Cyrillic, CJK, accented Latin letters, apostrophes, and mixed-language brand names more accurately in mention and position metrics.

Override a sentiment label, inspect its edit history, and revert the change on Growth and higher plans.

Use tables, tabs, date filters, and the assistant panel more comfortably on narrow screens.

See when a selected period has no data and whether a score covers the period or all tracked history.

Protect an account with authenticator codes and backup codes from Account Settings.

Add Copilot to an account and keep reports, exports, and emails limited to enabled models.

Follow long answers more easily with smart scrolling, clearer statuses, and revised chart styling.

Open Query Fan-Out from the main sidebar and see its data requirements when empty.

Understand unavailable sections through sidebar locks, upgrade guidance, and clearer empty states.

See when a new project is collecting its first data and refresh as soon as results arrive.

Create AI Reputation reports on Scale and Enterprise plans.

Use Cmd+K to find and open any product section available to the account.

Collect related pages and review their results as one group.

Compare visibility, traffic, sources, models, and prompts before and after a GEO change.

Jump from sentiment quick filters to the matching source pages and brand mentions.

Compare daily analytics totals with the share attributed to AI sources.

Ask the assistant for product guidance and see its conversation limit clearly.

Create prompts through clearer settings steps while preserving each suggestion's source and status.

Review a product feed's structure and page readiness for ChatGPT Shopping.

Import quoted values correctly from comma-separated tag files.

Read AI Visibility Score as a daily average and use the full selected period in Sentiment and brand filters.

Track more prompts across every self-serve plan.

Carry supported date, country, language, and other filters through sidebar navigation.

Keep a past-due subscription notice visible until the account issue is resolved.

Let selected team members create new projects without granting broader account control.

See which sign-in method you used last time directly on the login page.

Compare AI Visibility Score by competitor in AI Model Insights.

Open a response citation directly from the control beside Copy.

Filter Citation Sources by URL error status.

Save a preferred billing currency and see supported conversions throughout the account.

Find public help articles with search and category filtering.

Include the first observed date for each result in citation exports.

Share generated GEO tool reports with the whole account instead of only their creator.

Start projects through a more guided flow with richer prompt and competitor suggestions.

Manage supported payment methods such as SEPA and PayPal without billing assuming a card.

Find opportunities that combine citation and mention evidence with marketplace listings.

Review marketplace availability, authority metrics, purchase links, and local-currency prices with citation details.

Generate large Magic Prompt sets in the background without keeping the page open.

Identify and filter the background sources used in ChatGPT responses.

Use AI Traffic on Growth and higher plans, with Search Console prompt generation available on every plan.

See when each cited page first appeared in LLM Pulse tracking.

Compare your brand with any tracked competitor in AI Model Insights. Results keep zero-mention rows visible and include an Average column, position distribution, and the option to use your brand on either side.

Read citation rate and AI Visibility Score on dashboard and prompt cards with clearer empty states.

Explore AI Overview results with source-type filters and competitor selection in AI Model Insights.

Keep a selected custom range while moving between analytics pages.

Sign in with Google from the authentication page or Google One Tap.

Generate dedicated Social and UGC recommendations with their own settings and report flow.

Manage subscription status, payment methods, invoices, usage, and upgrades from a clearer workspace.

Carry the active date range into exports across product reports.

Analyze, filter, and export response and sentiment results by tag.

Share a generated report through a prominent public URL and clearer results page.

Follow a dismissible checklist for the remaining setup actions in Project Overview.

Use statuses and a clearer regeneration flow to manage recommendations as an action list.

Star frequently used projects so they remain easy to reach from the dashboard.

Choose which projects send weekly visibility summaries and negative sentiment alerts.

Use Gap Analysis and Sentiment views to find missing citations and negative pages more quickly.

Read each KPI with the change from the previous matching period.

Include screenshots and other images when opening a support conversation.

Receive a weekly email covering visibility, mentions, Share of Voice, sentiment, citations, and model performance.

See which themes drive brand discussion in a visual topic breakdown.

Pause projects directly and invite unlimited team members.

Decide which teammates can open Billing and Plans or the billing portal.

Track GPT-5.1 with or without web search alongside the existing AI models.

Filter prompts by tag, check the next scheduled update, and compare Citation Rate by Domain.

Find the right project without scrolling through a long account list.

Choose colors for competitors and keep them consistent across reports.

Collapse report and settings groups to reduce navigation clutter in larger accounts.

Upload and review a batch of tags before adding them to a project.

Update tag descriptions after creation to keep prompt collections understandable as projects grow.

Set the language used for sentiment extraction and recalculation in Project Settings.

Move between answers with Previous and Next controls and copy every citation in one action.

Copy cited URLs in one action and keep citation context visible even when an answer has no sources.

Keep working while accepted suggested prompts are added in the background.

Compare where your brand and competitors appear within each AI response.

Try LLM Pulse before subscribing, with clear trial status and plan guidance throughout the product.

Use the same searchable filter pattern across Overview, Prompts, Responses, Sentiment, and Citations.

Select an entire prompt table for bulk management without checking every row.

Review citation performance in one consistent Overview section instead of competing summaries.

Export the data behind Overview charts for use in reports and further analysis.

Search projects from the dashboard and see tracked competitor context on each project card.

Track AI Mode in Hindi, Indonesian, Japanese, Korean, and Portuguese as well as English.

Review and regenerate suggested prompts from the main prompt workflow in the project language.

Review the answer body, citations, and brand mentions in a clearer response detail view.

Limit invited teammates to selected client projects instead of sharing the entire account.

Assign prompts to a tag while adding them in bulk instead of cleaning them up later.

Model results, citations, and prompt context are easier to scan on the prompt detail page.

Account owners can move a project to another user from within the product.

Turn on brand highlighting to make mentions stand out while reviewing long AI responses.

New customers can create an account and start tracking their brand directly from the public site.

Regenerate suggested prompts without restarting project setup, then add the full set in one action.