MoltMentions
Book a demoSign up
‹
01 / 10AI answers
›
Features

Every AI visibility tracker feature in one API:
prompt panels, metrics, ChatGPT Ads, page audits and MCP.

Each feature below shows sample output and the call behind it, from AI visibility tracking to AI share of voice. Use them from code, from Claude or ChatGPT through the MCP server, or in the Monitor app.

AI answers

Track AI visibility across ChatGPT, Google AI Mode, AI Overviews, Gemini, Claude and Perplexity.

One API. One response shape. Brands named, sources cited, date.

  • ChatGPT
  • ChatGPT Ads
  • Gemini
  • Claude
  • Perplexity
  • Google AI Overview
  • Google AI Mode
  • Copilot
See the data for my brand
Prompt panels

Prompt tracking with prompt panels you create, store and run by tag.

Create prompts from tags, store them once, and run only the slice you need.

  1. 1
    CreatePrompts from your tags, in any market and language.
  2. 2
    StoreStorage is free, and you can query every tag.
  3. 3
    Run any slicePick tags and engines; only the matching prompts run.

Sample panel, made-up brand. Showing 4 of 25 prompts.

Create my first prompt panel
Stored answers

Store every AI answer once:
query it by brand, prompt, platform, source or date.

Your own record of what AI said, week after week. Keep answers 1, 3, 6 or 12 months; retrieving one costs $0.00005.

Query stored answers
# Illustrative sample.
GET /v1/answers?brand=northwind.com
    &platform=chatgpt
    &from=2026-09-01&group_by=week
{ "week": "2026-W44",
  "share_of_answers": 0.33,
  "top_sources": ["g2.com", "reddit.com"] }

Price per stored answer

Kept forPer answer3,000 answers
1 month$0.001$3.00
3 months$0.0012$3.60
6 months$0.0015$4.50
12 months$0.0021$6.30

Each extra month adds 10% of the first-month price. The API and the MCP server are included; Monitor adds the dashboard.

Store my first answers
AI visibility metrics

Stored answers automatically calculate your share of voice and citations in real time.

One snapshot shows where you are. A weekly record shows where you are heading. Share of voice, rank, mentions and citations on every stored answer: $0.005 an answer, included in Monitor.

Sample:
Northwind's share of AI answers rose from 8% to 33% in ten weeks, passing Contoso in week eight

0%25%50%W1W3W5W7W9W10Northwind 33%Contoso 20%Fabrikam 19%

Share of tracked answers naming each brand, by week. Sample data, made-up brands.

Start my weekly record
ChatGPT Ads monitoring

ChatGPT Ads monitoring shows which advertisers appear on your prompts.

See which rival pays to appear on your prompts, with what ad, on which day. Live now.

Ads seen on your prompts

PromptAdvertiserAd textSeen
best crm for small teamsLeadLarkClose more deals this quarterNov 5
crm with email syncLeadLarkCRM that fills itself inNov 5
sales tool for agenciesFabrikamWin clients with proofNov 4
crm pricing comparedSendloopNo ad seenNov 4

Sample data, made-up advertisers.

Sample:
LeadLark ramped its ChatGPT ads from 6% to 47% of prompts in six weeks, while Sendloop pulled back to 0%

0%25%50%W1W3W5W6LeadLark 47%Fabrikam 18%Sendloop 0%

Share of tracked prompts showing each advertiser's ad, by week. Sample data, made-up advertisers.

Watch ads on my prompts
Page audit

One page audit call scores a page’s AI readiness.

Send a URL and get back a score out of 100, the fixes to make, and whether AI engines cite the page today. Live now.

AI readiness score62 of 100

AI bots can crawl it10
Answer in first lines6
Clear headings8
Schema markup3
Facts and sources5
Fresh date9
Fast, readable page7
Cited by AI today2

Sample result for one made-up page. Each check scores 0 to 10.

Audit a page
# Audit one URL. Add a prompt to test if AI cites it.
POST /v1/audit/page
{ "url": "https://northwind.com/pricing",
  "prompt": "best project tool for agencies" }

# Response
{ "score": 62,
  "fixes": [
    { "check": "schema", "do": "Add FAQPage markup" },
    { "check": "answer_first", "do": "Lead with the price" } ],
  "cited_by": ["perplexity"] }
Audit my first page
Recommendations

AEO and GEO recommendations:
what to fix on your pages, and where to get cited.

Each recommendation reads what AI said about you and what your pages show, and returns a ranked list of fixes.

Free with a page audit

Content and technical

  • AEOWhat to fix to get named in answers
  • SEOPages and queries to improve
  • CrawlersWhat AI bots can and cannot reach
  • Entity schemaMarkup that tells AI who you are

Included with every page audit ($0.23 a page).

10 credits per stored answer

Owned and earned media

  • Owned mediaContent to publish on your own site
  • Earned mediaSources that should mention you
  1. Publish a client reporting template page
  2. Get listed in two agency roundups on g2.com
  3. Answer pricing in the first lines of /pricing

Worked out from the citation gaps in your stored answers: 3,000 answers cost about $12.21. Sample fixes, made-up brand.

Response: content and technical
# Illustrative sample.
GET /v1/audits/aud_7f3k/recommendations
{ "url": "northwind.com/client-reports",
  "ai_readiness": 62,
  "cost": "free with the audit",
  "recommendations": [
    { "area": "crawlers", "priority": 1,
      "fix": "Allow GPTBot in robots.txt" },
    { "area": "entity_schema", "priority": 2,
      "fix": "Add FAQPage and Organization JSON-LD" },
    { "area": "aeo", "priority": 3,
      "fix": "Answer pricing in the first two lines" }
  ] }
Response: owned and earned media (10 credits per stored answer)
# Illustrative sample.
POST /v1/recommendations
{ "panel_id": "pnl_agency_us", "media": ["owned", "earned"] }
{ "answers_read": 75, "credits_used": 750,
  "citation_gaps": [
    { "domain": "g2.com", "cites_rivals": 18, "cites_you": 0 },
    { "domain": "reddit.com", "cites_rivals": 12, "cites_you": 2 }
  ],
  "recommendations": [
    { "media": "earned", "priority": 1,
      "fix": "Get listed on g2.com" },
    { "media": "owned", "priority": 2,
      "fix": "Publish a client reporting template page" }
  ] }

Sample responses, made-up brand.

Get my first recommendations
Attribution tracking

AI traffic attribution:
tie AI visits, sign-ups and revenue to the answers that sent them.

See whether AI search makes you money, in three steps.

  1. 1

    Connect your sources

    • GA4 or ThriveStack Marketing Analyticswebsite visits and where they came from
    • Server logsAI crawler hits on your pages
    • Stripepayments
    • CRMleads and deals
  2. 2

    Match the journey

    Each visit from an AI assistant is matched to the sign-up, customer and payment it led to.

    ChatGPT answer→Visit→Sign-up→$ Revenue
  3. 3

    Read the report

    Visits, AI crawler hits and revenue by month, so you can see whether AI search is making money.

    See the sample report ↓
Traffic and revenue by monthSample report
0$0K500$10K1,000$20K1,500$30K2,000$40KMayJunJulAugSepOct12K61K$38K
DirectOrganicAI assistantsAI crawler hitsRevenue

How to read this chart

Hover or tap a month to see its numbers.

  • Bars: visits each month, stacked: direct and organic in grey, AI assistants in orange.
  • Dashed line: how often AI crawlers read your site. More crawling usually comes before more AI visits.
  • Green line: revenue from those visits, on the right-hand axis.
  • Revenue lags traffic: people sign up and pay days or weeks after their first visit, so revenue climbs about a month behind AI visits: July's jump in AI visits shows up as August's jump in revenue.
  • What this sample says: AI assistant visits grew from 30 to 740 a month while direct and organic grew slowly, and revenue rose with them, from $5K to $38K.

An add-on for Monitor: $20 per 100,000 events a month, billed in 100,000-event blocks. Sample data, made-up brand.

Track my AI revenue
MCP

The MCP server puts AI visibility data inside your AI agent.

54 tools, 53 read only. Sign in with OAuth or a token. Setup docs

# Settings, Connectors, Add custom connector. Sign in, no key.
https://api.app.thrivestack.ai/mcp
# Then run /mcp in a session to sign in.
claude mcp add --transport http --scope user Thrivestack \
  "https://api.app.thrivestack.ai/mcp"
# Cursor settings, MCP. Click Connect to sign in.
{ "mcpServers": { "thrivestack": {
    "url": "https://api.app.thrivestack.ai/mcp" } } }
# For clients without OAuth. Create a token in Settings.
https://api.app.thrivestack.ai/mcp
Authorization: Bearer YOUR_PAT_TOKEN
Which sources does ChatGPT cite for "best project tool for a small agency"?
Agent called get_prompt_sourcesChatGPT cites three sources most often: g2.com, northwind.com and reddit.com. Your brand is named in 2 of 3 answers.

Sample data, made-up brand.

Connect my agent
FAQ

AI visibility features, explained.

What features does MoltMentions have?

AI answers from ChatGPT, Google AI Mode, AI Overviews, Gemini and more in one response shape; prompt panels you create from tags; stored answers you can query; AI visibility metrics such as share of voice and citations; ChatGPT Ads monitoring; page audits for AI readiness; content, technical, owned and earned media recommendations; attribution that ties AI visits to revenue; and an MCP server so Claude, ChatGPT or any MCP client can use all of it.

Which MoltMentions features are live today?

All of them: AI answers from every engine, prompt panels, stored answers, AI visibility metrics, ChatGPT Ads monitoring, page audits, recommendations, attribution tracking and the MCP server.

Do I need the Monitor plan to use these features?

No. On the Wallet you pay per request from credits and call each feature through the API or the MCP server. Monitor adds the schedule, the app and AI visibility metrics on every stored answer, priced by prompt volume.

Can I use MoltMentions features from Claude or ChatGPT?

Yes. The hosted MCP server has 54 tools, and 53 of them only read data. Connect it to Claude, Claude Code, Cursor or any MCP client and ask for runs, metrics, audits or recommendations in plain language.

Get started

Try every feature from $5.

Pay as you go, no monthly fee. Or pick Monitor and get the app, the schedule and the metrics in one plan. More studies in ThriveStack research.

Sign upSee pricing