Skip to main content
Cheyenne
mEmployee
CheyennemEmployee
mEmployee
June 25, 2026

Meltwater GenAI Lens

  • June 25, 2026
  • 0 Replies
  • 47 views

What is GenAI Lens?

GenAI Lens is Meltwater's AI-generated content monitoring tool that shows you how your brand, products, competitors, and topics appear across the world's leading AI assistants — including ChatGPT, Gemini, Perplexity, Grok, Claude, and others. It gives you visibility into what large language models (LLMs) say about you, which sources shape those answers, and how narratives shift over time.

 

With AI tools now serving as the first stop for many consumers researching brands and products, GenAI Lens lets you see this layer early — so you can influence it and protect your reputation.

 

What you can do with GenAI Lens:

  • Monitor how your brand, products, and competitors appear in AI-generated responses
  • Track sentiment, emotions, and key phrases tied to AI responses about your brand
  • Identify the journalists, outlets, and sources that influence LLM answers
  • Detect inaccuracies, negative themes, or misinformation early
  • Compare your presence and share of voice against competitors
  • Generate LLM Executive Reports for leadership briefings

 

How to access: Click the Monitor drop-down in the left-hand navigation bar and select GenAI Lens.

 

Subscription note: The number of saved prompts available to you depends on your Meltwater subscription. Check the top of the prompt creation workflow screen to see your total prompt allowance and how many you have used. To add more prompts, contact your Account Team.

 

Key concepts

  • Prompt A question or instruction sent to one or more LLM models to retrieve how they respond about your brand, topic, or competitor. Prompts are the foundational unit of monitoring in GenAI Lens.
  • Prompt Template A prebuilt prompt organized by use case — such as Brand Monitoring, Competitive Intelligence, or Crisis Management — available in the prompt creation workflow. Templates speed up setup and are designed around common research objectives.
  • Prompt Variation An automatically generated alternate version of your prompt. GenAI Lens creates up to 10 variations to increase depth and accuracy of insights. Each variation counts toward your total prompt limit.
  • Prompt Topic An organizational folder that groups related prompts. Topics can be structured by sub-brand, product, competitor, strategic initiative, or market area.
  • LLM / AI Assistant Model The AI assistant whose responses are being monitored. GenAI Lens currently supports more than 90% of widely used LLMs, including:

    • Anthropic Claude Opus 4.1

    • Anthropic Claude Opus 4

    • ChatGPT 5

    • DeepSeek R1

    • Google Gemini 2.5 Flash

    • Google Gemini 2.5 Pro

    • Meta AI Llama 4 Maverick

    • Perplexity Sonar Pro

    • xAI Grok 3

    • xAI Grok 4

  • Visibility Score A composite metric combining Mention/Citation Rate (reach), Frequency (intensity), and Prominence (placement in the text) into a single number. Higher scores indicate the entity is widely mentioned, appears often, and is featured early in LLM responses.
  • Mention Rate The percentage of LLM results that contain a reference to a given entity. Measures the reach of an entity across your dataset.
  • Prominence Score Measures how early an entity appears within a prompt's response. Scores range from 0 to 100, where 100 means the entity appears at the very beginning and 0 means it appears at the very end.
  • AI Recommendations AI-generated strategic recommendations that appear on the Overview tab once at least one brand is set up and related prompts are selected.
  • Brand Setup The process of registering your brand (or a competitor's brand) in GenAI Lens by adding known aliases, links, people, and products. Brand setup enables data to appear in the Benchmarking and Trends tabs.
  • Target Country A geographic location applied to a prompt so that responses reflect localized search behavior. Prompts without a target country return globally scoped responses and cannot be filtered by region.

 

Frequently asked questions

What AI models does GenAI Lens support?

GenAI Lens currently supports more than 90% of widely used LLMs, including Anthropic Claude Opus 4.1, Anthropic Claude Opus 4, ChatGPT 5, DeepSeek R1, Google Gemini 2.5 Flash, Google Gemini 2.5 Pro, Meta AI Llama 4 Maverick, Perplexity Sonar Pro, xAI Grok 3, and xAI Grok 4. Meltwater is actively expanding coverage as new models enter the market. Note: ChatGPT, Google AI Overviews, Google AI Mode, DeepSeek, and Perplexity are included with GenAI Lens Lite plans.

 

How many prompts can I create?

The number of saved prompts depends on your Meltwater subscription. Your total allowance and current usage are displayed at the top of the prompt creation workflow screen. To add more prompts, contact your Account Team.

 

How quickly do results appear after I create a prompt?

Results take approximately 10 to 20 minutes to appear on the Overview page after prompt creation, as GenAI Lens enriches the data. Results then update every 24 to 48 hours depending on your package. If you are unsure which package you have, contact your Account Team.

 

What are Prompt Variations and do they count against my limit?

Prompt Variations are automatically generated alternate phrasings of your prompt. They increase the depth and accuracy of insights by capturing different ways users might phrase the same question. Each variation counts toward your total prompt allowance. You can select up to 10 variations per prompt, preview each one, or edit them before saving.

 

What does a Target Country do?

Adding a target country to a prompt causes the system to assign an IP address in that region when querying LLMs, returning more localized results. Without a target country, responses are globally scoped and cannot be filtered by region. For example, a prompt about "the best football team" returns different results depending on whether the UK or US is selected, as it affects which sport's teams the LLM references.

 

Why does prompt language matter?

The language used to write your prompt influences which sources the LLM cites. Writing a prompt in French while targeting France, for instance, is likely to surface more French-language sources than writing the same prompt in English.

 

Why are there no citation or citation rate metrics for Meta Llama or DeepSeek?

DeepSeek and Meta Llama do not have web search capabilities, so they do not provide citations in their responses. Citation-related metrics are therefore unavailable for these models in the Benchmarking tab.

 

What is the difference between the Overview tab, the Trends tab, and the Benchmarking tab?

Overview: Displays individual LLM responses in a grid view, with columns for model, prompt, date, emotion, key phrases, brands, products/people, and links. Click any row to open a detailed prompt analysis panel.

Trends: Shows trend graphs, a ranking grid, and a word cloud across seven categories — Brands, Products, People, Locations, Key Phrases, Sources, and Links — with toggles for competitor-only mode and AI Insights.

Benchmarking: Provides heat maps and ranking widgets comparing entities across visibility metrics. Uses AI Insights to summarize the data. Requires brand setup to be completed first.

 

How do I export my results?

From the Overview tab, click Download and choose Excel, JSON, or PNG. The exported Excel file includes two tabs: an Overview tab (summary of date range, selected models, and prompts) and a Results tab (detailed output per prompt, including sentiment, key phrases, organizations, products, people, and cited links). You can also export a single prompt's analysis as a CSV from the prompt analysis slide-out panel in the results grid.

 

How do I generate an LLM Executive Report?

Click Generate Executive Report in the top-right corner of GenAI Lens, enter a report name, apply filters (date, models, prompts, labels, countries), and click Generate Report. Reports typically take approximately 20 minutes to complete and download as a PowerPoint file. During the beta period, reports are manual-only and cannot be automated or scheduled. Use the View Past Reports option to check status and download completed reports.

 

How-to guides

How to create a prompt

Use this when: You want to begin monitoring how your brand, a competitor, or a topic appears in AI-generated responses.

 

Before you begin: Spend a few minutes researching your core business priorities, reviewing your existing Explore searches, and considering how different audiences phrase questions — consumers use shortcuts, journalists add context, risk managers write formally. Strong results depend on thoughtful prompts.

 

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Click Settings
  4. Click Create a prompt. The total number of prompts available in your account and how many you have used are displayed at the top of the workflow screen.
  5. Choose a template or select Custom Prompt. Templates are organized by use case tabs: Brand Monitoring, Competitive Intelligence, Industry Trends, Crisis Management, and others.
  6. Customize your prompt:
    • Update the Prompt Name if needed.
    • Add a Target Country using the drop-down if you want localized results.
    • In the Prompt field, replace the bracketed placeholders with your brand, competitor, or topic information — or write your own prompt if you selected Custom Prompt.
    • Add a Label using the Labels drop-down.
    • Select a Prompt Topic to organize the prompt. You can also create a new topic here. Topics are recommended to be structured by sub-brand, product, competitor, strategic initiative, or market area.
  7. Click Preview Prompt to see a sample of the type of data this prompt would generate in the right-hand slide-out.
  8. For further edits, click Edit Prompt. When satisfied, select Looks Good.
  9. Review Prompt Variations. GenAI Lens automatically generates up to 10 variations to increase depth and accuracy. Use the eye icon to preview a variation or the pencil icon to edit it. Each variation counts toward your total prompt limit.
  10. Check the box next to each variation you want to include.
  11. Click Create Prompt.

 

Note: Data collection begins once the prompt is created. Results take 10 to 20 minutes to appear on the Overview page and update every 24 to 48 hours depending on your package. If you are unsure when a prompt was created, check the Settings tab.

 

Result: Your prompt is saved and begins collecting data. Results appear on the Overview page within approximately 10 to 20 minutes.

 

How to manage and organize prompts

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Click GenAI Lens.
  3. Select Settings in the top toolbar.
  4. To take bulk actions: select the checkboxes to the left of two or more prompts, then click Move to topic to assign them to a topic, or click Delete to remove them.
  5. To take actions on a single prompt: click the Topic drop-down to move the prompt to a topic, or click the kebab menu to edit the prompt name or delete it.
  6. To take action on an entire topic: navigate to the Prompts section on the left side of the screen, hover over the topic, and click Edit or Delete.
  7. To create a new topic: click Create Topic, enter a topic name, and click Create.

 

Result: Your prompts are organized into topics for streamlined navigation and reporting.

 

How to bulk upload prompts

Use this when: You need to add a large number of prompts at once and prefer to work from an Excel template.

 

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Click Settings in the top toolbar.
  4. Select Bulk Add Prompts.
  5. Click Download Excel Template.
  6. Open the template and complete the following columns:
    • Name: A recognizable label for the prompt. Required. If missing, the system will display an error: "Invalid file format. Missing required columns: Prompt Name. Please use the template file."
    • Prompts: The exact text sent to the LLMs. Required.
    • Topics: Automatically assigns the prompt to an existing topic or creates a new one if the name does not match an existing topic.
  7. Return to Meltwater and upload your completed file.
  8. Click Next.
  9. Review the status icons in the Review phase before finalizing:
    • Valid: Prompts that will be successfully uploaded.
    • Errors: Prompts with issues. Common error types include:
      • "You have reached the account prompt limits for your company" — delete prompts or purchase more.
      • "Failed to create prompt" — ensure both the name and prompt fields are filled in.
      •  "Internal_Server_Error" — re-upload the prompt(s). If the issue persists, file a support ticket
    • Unable to upload – limit reached: Prompts that will not be uploaded because the prompt limit has been reached.
  10. Once finished, click Add Prompts.

 

Note: You can add more prompts for an additional cost, or delete existing ones from the list. To add more prompts, connect with your Account Team.

 

Result: Your prompts are uploaded and processed. The Review page confirms which prompts uploaded successfully (Valid) and which encountered errors.

 

How to view and analyze results in the Overview tab

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Results are displayed in Grid View by default.
  4. The results grid has eight columns: Model, Prompt, Date, Sentiment, Key Phrases, Brands, Product/People, and Links.
  5. Click any row to open the prompt analysis slide-out panel, which shows: AI model and last data retrieval time, prompt name, prompt text, verbatim LLM response (with copy option), top emotion, top key phrases, top brands, top products and people, and top links with clickable outbound links.
  6. The slide-out panel also includes an Export button that downloads all prompt analysis data as a CSV file.

 

Result: You can review individual LLM responses, compare results across models, and export detailed analysis data.

 

How to filter results

  1. From the GenAI Lens results page, use the filters at the top of the screen:
    1. Date range: Filter to a specific date or date range. Responses from all models are collected daily. Each response includes a collection date shown in the Day column.
    2. Saved prompt: Select any combination of prompts, or delete prompts from the dropdown.
    3. LLM/AI Assistant Model: Filter by one or more specific models.
    4. Brand: Filter to results associated with a specific brand you have set up in Brand Setup.

 

Result: The results grid and word cloud both update dynamically to reflect the applied filters.

 

How to explore the Trends tab

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Click Trends.
  4. Optionally, toggle Competitor-only mode on or off to filter results to your brand alongside selected competitors.
  5. AI Insights at the top of the tab summarize the data. Click Regenerate Summary to refresh, or Copy To Clipboard to copy the summary.
  6. The Trends tab has three visualizations, each supporting seven categories — Brands, Products, People, Locations, Key Phrases, Sources, and Links:
    • Trends Graph: Displays trends over time for the selected category. Hover over any day to see the data for that date. Click legend items to show or hide them. Use the Mentions or Citations drop-down to toggle between Mentions, Mention Rate, or Visibility.
    • Word Cloud: Shows entities from the results. Hover over any word to see how many times it was mentioned. Covers Key Phrases, Organization/Brand, Product, Location, and People.

 

Result: You can track how entities and themes trend over time and compare relative visibility across brands, products, sources, and more.

 

How to use the Benchmarking tab

Before you begin: Ensure at least one brand is set up in Brand Setup and that related prompts have been selected. The Benchmarking tab requires brand setup to populate data. See the Setting Up Your Brand how-to guide below.

 

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Click GenAI Lens.
  3. Select Benchmarking in the top toolbar.
  4. In any graph, click the Chart Type drop-down to switch between Pie Chart, Bar Chart, and Table views.

 

Note: Most heat map data is provided on a scale of 0 to 100. When using Total Mentions, the data compiles the entity's total across various LLMs. Sentiment scoring is assigned by a system prompt and uses the LLMs to give a score — for both the entire result and per entity.

 

Result: You can compare entities across visibility metrics using heat maps and ranking widgets, with AI Insights summarizing the data.

 

How to set up your brand

Use this when: You want your brand or a competitor to appear in the Benchmarking and Trends tabs, or to enable AI Recommendations on the Overview tab.

 

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Click Brand.
  4. Select Add Brand.
  5. Type in the brand name you wish to add.
  6. Add known Aliases, Links, People, and Products to associate with the brand filter: click the drop-down to the right of each category, enter what you would like to add, and click Add.
  7. To edit: click the pencil icon. To delete: click the X.
  8. Click Next.
  9. Enter your Competitors in the text box. Add or remove aliases, links, products, and people for competitors using the pencil icon.
  10. Click Continue to Add Prompts to finish Brand Setup.

 

Note: Brand Setup applies to the Benchmarking and Trends tabs. Under the Sources tab, two new filter options will appear once a brand is set up: one for competitors and one for owned sources.

 

Result: Your brand is registered in GenAI Lens. Data will begin populating in the Benchmarking and Trends tabs, and AI Recommendations will appear on the Overview tab.

 

How to use chart drill-downs

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Click on any entity in a results chart or any point in a graph to access drill-downs.
  4. In the right-hand slide-out, review the Results tab to see which LLM responses mention the selected entity or data point.
  5. Click the Analytics tab to see visibility by model and the top 10 prompts powering the results.

 

Result: You can drill into specific data points to see the underlying LLM responses and understand which prompts and models drive a given trend or result.

 

How to add a location to a prompt

A target country can be added during prompt creation or later through Settings.

 

During prompt creation:

  1. In Step 2 of the prompt creation workflow, after adding your prompt name, click the Target Country drop-down.
  2. Select a country.
  3. Complete the remaining required fields and click Create Prompt.

 

Through Settings after creation:

  1. Click Settings in the top toolbar. (Depending on your screen size, you may need to click the arrow to view remaining top tabs including Settings.)
  2. Click the kebab menu on the prompt you want to add a location to.
  3. Click Duplicate.
  4. Open the duplicated prompt.
  5. In Step 2, click the Target Country drop-down and select a country.
  6. Complete the remaining required fields and click Create Prompt.

 

Note: When no target country is assigned, the system returns a globally scoped response and you will not be able to drill down into results for a specific region. A location must be specified when prompts are sent to LLMs so that an IP address in that region can be assigned.

 

Result: The prompt queries the LLM using an IP address associated with the selected country, returning more localized responses.

 

How to generate an LLM Executive Report

  1. Click the Monitor drop-down in the left-hand navigation.
  2. Select GenAI Lens.
  3. Click Generate Executive Report in the top-right corner.
  4. Select Generate Executive Report.
  5. Enter a Report Name.
  6. Use the drop-downs to add relevant filters: Date, Models, Prompts, Labels, or Countries.
  7. Click Generate Report. A notification appears in the top-right corner confirming the report is being generated.
  8. To check the status, click the Generate Executive Report drop-down and select View Past Reports.
  9. Once your report is ready, click Download. The report downloads as a PowerPoint file, ready to upload to Google Slides or any presentation tool.

 

Pro tip: Before sharing with leadership, review the recommendations slide. The reasoning behind each recommendation is included in the report — knowing it upfront helps you field questions confidently in the room.

 

Note: Reports typically take approximately 20 minutes to complete. During the beta period, reports are manual-only — automation is not available. Pricing and packaging for additional reports are still being finalized. Feedback on reports can be submitted through mCommunity Ideas or when you have used all report credits.

 

Result: A branded PowerPoint report summarizing your LLM visibility, sentiment, competitive positioning, and strategic recommendations is downloaded and ready to share with leadership.

 

Related articles

      Terms & ConditionsAccessibility statement