# Lantad compared with Semrush

> Semrush's AI Visibility Toolkit reports how often AI answers mention your brand; Lantad measures whether an AI crawler could read the page those answers would have to cite, and names the fix when it could not.

- Canonical page: https://lantad.co/vs/semrush
- This file: https://lantad.co/vs/semrush.md

## Key facts

- **Compared with:** Semrush
- **What they do:** Semrush sells the AI Visibility Toolkit as a subscription alongside its SEO products. Its own knowledge base article, at semrush.com/kb/1493-ai-visibility-toolkit, names six components: Visibility Overview, Competitor Research, Prompt Research, Brand Performance, Prompt Tracking, and AI Search Site Audit. The headline metric is an AI Visibility Score, which Semrush defines as how often your brand is mentioned in AI-generated answers compared with the median number of mentions for your top industry competitors. Brand Performance analyses share of voice, brand sentiment, and, in Semrush's words, the key narratives driving your reputation, by running queries through LLMs.

Engine coverage on its pricing page, semrush.com/pricing/ai, is stated as mentions from ChatGPT, Google AI, Gemini and Perplexity. The Base plan is 99 USD a month per domain billed annually, and includes 25 custom prompts for daily tracking, one domain for Brand Performance analysis, 300 daily queries in AI Analysis reports, 1000 daily queries in Prompt Research, AI Search Checks in Site Audit for up to 100 pages, and 10 CSV exports a day. Semrush One, which bundles the SEO and AI Visibility toolkits, starts at 199 USD a month.

Semrush also ships a genuine crawler-side check, and this page does not pretend otherwise. Its AI Search Site Audit and the AI Search Health widget inside Site Audit flag technical issues that might prevent AI bots from crawling your content. That is a real overlap with what Lantad does, and it is the part of the comparison worth reading carefully rather than the part worth glossing over.
- **What Lantad does:** Lantad fetches a page twice. Once as a plain HTTP request identifying as LantadBot with no JavaScript, which is how most AI crawlers take a page, and once in a real browser. The difference between those two documents is content an AI crawler never received. That difference is scored as Prose Parity, which carries half the weight of the whole grade because it is the failure that hides: the page looks perfect to you and arrives empty to the crawler.

The rest of the grade is Access at 25 percent, Structure at 15 percent and Schema at 10 percent, and the methodology page publishes every weight and threshold rather than describing the score in adjectives. Access is checked per crawler, against robots.txt and against the live server response separately, across 15 named AI crawlers, because those two sources disagree more often than anyone expects. Where the scan finds a defect it names the remedy from a catalogue of 24, with stack-specific guidance for Next.js, React, Shopify, Framer and Lovable.

On the answer side Lantad is narrower than Semrush and says so. Paid plans track 100 to 250 prompts weekly across ten pages, against up to 8 engines: 5 metered API engines, an always-on open-weight chain, a grounded engine, and Google AI Overviews captured from the public SERP through a vendor rather than by automating a consumer app. The metered engines answer the top 10 prompts of each page run, which every run report discloses, so the engine count is never a prompts-times-engines multiplier. Sentiment is model-graded on paid plans with a labelled word-list fallback. There is no region or language dimension in a run. The numbers in this paragraph are read from the same configuration the scheduler enforces, not retyped into the page.
- **Choose them when:** Choose Semrush when you want one account and one invoice. If your team already runs keyword research, rank tracking and site audits there, the AI Visibility Toolkit adds the answer layer without a new vendor, a new login or a new procurement conversation, and that is worth more in practice than most feature comparisons admit. It is also the better choice when you need share of voice against an industry median, brand sentiment and narrative analysis, prompt discovery at 1000 queries a day, or daily cadence on the prompts you care about most. Lantad does none of those and is not trying to.
- **Choose Lantad when:** Choose Lantad when the question is whether the page was readable in the first place. A visibility score that says you are absent tells you the outcome and not the cause, and the most common cause is mechanical: body text that only exists after JavaScript runs, a robots.txt group that allows Googlebot and refuses GPTBot, or a bot rule at the CDN that contradicts the robots file. Lantad measures those directly per crawler, scores them, and names the fix, and the free scan needs no signup and no card, so you can settle the question before deciding whether to buy anything at all.

## Lantad against Semrush, dimension by dimension

| Dimension | Semrush | Lantad |
| --- | --- | --- |
| Primary question answered | How often do AI answers mention my brand, and how does that compare with my competitors | Did the AI crawler receive my page, and if not, which line of which file caused it |
| Core method | Runs queries through LLMs and counts mentions, citations, sentiment and share of voice | Fetches the page as a crawler and as a browser, diffs the two documents, and scores the gap |
| Crawler access checking | Partly: AI Search Site Audit and the AI Search Health widget flag issues that might block AI bots, on up to 100 pages on the Base plan | The whole product: robots.txt and the live server response checked separately per crawler across 15 named bots, then the raw-versus-rendered diff on top |
| Tracked prompts | 25 custom prompts checked daily on the Base plan, which is 175 checks a week across 25 distinct questions | 100 distinct prompts a week on Starter and 250 on Business, each sampled once weekly and twice weekly on Business |
| Engines named | ChatGPT, Google AI, Gemini and Perplexity | Up to 8: 5 metered API engines, an open-weight chain, a grounded engine, and Google AI Overviews via public-SERP capture |
| Fixing what it finds | Reports issues and opportunities; the remediation is yours to carry out | 24 catalogued defects each with a named remedy, plus stack guides and paid Deep Audit and Fix Sprint work |
| Published price | 99 USD a month per domain for the AI Visibility Toolkit, or from 199 USD a month for Semrush One with the SEO toolkit included | Free scan with no signup, then published plans at 59, 199 and 399 USD a month, none of which can be bought yet: Lantad is pre-revenue and the pricing page says so |
| Breadth of the suite | Wide: AI visibility sits beside keyword research, backlinks, rank tracking, site audit and content tooling in one account | Narrow on purpose: machine readability, with a small answer-side surface attached to it |

## Which to choose

| Choose | When |
| --- | --- |
| Semrush | Choose Semrush when you want one account and one invoice. If your team already runs keyword research, rank tracking and site audits there, the AI Visibility Toolkit adds the answer layer without a new vendor, a new login or a new procurement conversation, and that is worth more in practice than most feature comparisons admit. It is also the better choice when you need share of voice against an industry median, brand sentiment and narrative analysis, prompt discovery at 1000 queries a day, or daily cadence on the prompts you care about most. Lantad does none of those and is not trying to. |
| Lantad | Choose Lantad when the question is whether the page was readable in the first place. A visibility score that says you are absent tells you the outcome and not the cause, and the most common cause is mechanical: body text that only exists after JavaScript runs, a robots.txt group that allows Googlebot and refuses GPTBot, or a bot rule at the CDN that contradicts the robots file. Lantad measures those directly per crawler, scores them, and names the fix, and the free scan needs no signup and no card, so you can settle the question before deciding whether to buy anything at all. |

## The short version

Semrush is the tool most people in this comparison are already paying for, which changes the question. Nobody lands here wondering whether Semrush is good software. They land here having seen the AI Visibility Toolkit appear inside an account they already have, and wanting to know whether it covers the job, or whether something is still missing.

The honest answer is that it covers most of the answer layer and part of the page layer, and that the part it does not cover is the one that decides whether anything else can work. Semrush measures how often AI answers name you. Lantad measures whether the crawler that feeds those answers received your page at all. This page sets out both, using Semrush's own published figures, and says plainly where Semrush is the better buy.

## Why a rank tracker cannot see this failure

The structural point is worth stating plainly, because it is the whole reason two tools exist rather than one. Ranking is decided after a page has been fetched, parsed and indexed. Citation inside an AI answer is decided during retrieval, while the model or its retrieval layer is assembling context. Those are different moments, and a page can pass the second test only if it survived the first.

That ordering has a consequence most SEO dashboards are structurally unable to show. A rank tracker reports on pages that were successfully crawled, because a page that was never crawled has no ranking to report. The failure mode Lantad exists for is invisible to it by construction: the page ranks, the report is green, and the crawler that would have cited it received an empty shell. This is not a criticism of Semrush's engineering. It is a property of what a ranking metric is.

If the vocabulary here is unfamiliar, [answer engine optimization](https://lantad.co/glossary/aeo) sets out the whole chain from crawl to citation, and [generative engine optimization](https://lantad.co/glossary/geo) covers the rival name for the same category. [AI visibility](https://lantad.co/glossary/ai-visibility) is the term both vendors use for the outcome being measured.

## What the AI Search Site Audit does, and where it stops

Semrush ships a crawler check, and any comparison that ignored it would be dishonest. The AI Search Site Audit and the AI Search Health widget inside Site Audit, both documented at semrush.com/kb/1493-ai-visibility-toolkit, flag technical issues that might prevent AI bots from crawling your content, across up to 100 pages on the Base plan. If your robots.txt blocks an AI crawler outright, that is the kind of thing a health widget is built to surface, and it will surface it.

The difference is what happens when the robots file and the server disagree. Robots groups are matched per product token, which is what [RFC 9309](https://datatracker.ietf.org/doc/html/rfc9309) specifies and why [OpenAI's GPTBot](https://platform.openai.com/docs/bots) and [Google's Google-Extended](https://developers.google.com/search/docs/crawling-indexing/overview-google-crawlers) have to be named separately rather than covered by one wildcard. A robots.txt that allows GPTBot is a statement of intent; what the crawler actually receives is a fact, and the two part company whenever a WAF rule, a bot-management setting or a geo rule sits in front of the origin. Lantad checks both sides separately for each of [15 named crawlers](https://lantad.co/tools/ai-crawlers) and reports the disagreement as its own finding, because the disagreement is the bug. [The robots.txt tester](https://lantad.co/tools/robots-txt-tester) shows the first half and [what GPTBot sees](https://lantad.co/tools/what-gptbot-sees) shows the second, both free and both without an account.

The second difference is the diff itself. A health check asks whether the crawler was allowed in. Lantad asks what it came away with, by fetching the page without JavaScript and comparing that document with the rendered one word by word. A page can be fully allowed, return a clean 200, and still hand a crawler a navigation bar and a spinner. That is not a permissions problem and no permissions check will find it.

## Twenty five prompts daily against a hundred weekly

The prompt allowances look comparable and are not, and the difference runs in both directions, so it is worth being precise rather than selective.

Semrush's Base plan tracks 25 custom prompts with daily rankings. Over a week that is 175 checks, spread across 25 distinct questions. Lantad's Starter plan tracks 100 distinct prompts weekly, Business 250 with twice-weekly sampling. So Lantad watches four to ten times as many distinct questions, and Semrush watches its 25 seven times as often.

Which is better depends entirely on what you are trying to see. Daily cadence catches movement: if you want to know that an answer changed this Tuesday, a weekly sample cannot tell you. Breadth catches coverage: if you want to know which of a hundred buying questions you are absent from, 25 slots will not hold them. Neither allowance is generous enough to do both, and a vendor claiming otherwise is describing a bigger bill.

One more thing worth knowing about any prompt number, ours included: sampled answers are not measurements. An engine asked the same question twice can answer differently, so a share figure from a small sample carries real uncertainty. Lantad states the sample size beside the number rather than presenting a percentage as though it were counted, and [what AI says](https://lantad.co/what-ai-says) shows the shape of that output before you pay for anything.

## What each one costs, and one thing to know about ours

Semrush publishes 99 USD a month per domain for the AI Visibility Toolkit billed annually, and from 199 USD a month for Semrush One, which bundles it with the SEO toolkit. If you already hold a Semrush subscription, the marginal cost of the AI layer is the honest number to compare against, and it is the strongest argument in Semrush's favour.

Lantad's scan is free, needs no signup, and returns the full report and the fix list before any form. The published plans are 59, 199 and 399 USD a month. None of them can be bought today. Lantad is pre-revenue, [the pricing page](https://lantad.co/pricing) says exactly that, and this page is not going to imply otherwise in order to look like a finished business. What is live and working is the scanner, the report, the fix list and the free tools.

The comparison that actually matters is not price against price. It is whether the cheaper question has been answered first. Paying either vendor to monitor answers about a site whose pages arrive empty to the crawler buys a dashboard that faithfully reports zero.

## Running both, in the order that works

Most teams that end up on this page should run both tools, and the order is not arbitrary.

Start with readability, because it is the cheapest to test and the only layer where a fix has a mechanical effect. [Scan the page](https://lantad.co/) free, read the Prose Parity number, and check the crawler access matrix for the bots you care about. If body text is missing without JavaScript, that is the finding, and [the fix list](https://lantad.co/fix) names the remedy for the stack you are on.

Fix what the scan found, then re-scan the same URL to confirm the crawler now receives the text. This is the one before-and-after in the whole chain that is genuinely measurable, because both sides of it are things you can fetch.

Then let the monitoring run. Semrush's daily cadence on 25 prompts is well suited to the period after a fix, when you want to see whether anything moves. What nobody can promise, us included, is that it will: citation is an outcome decided inside systems no vendor can read. Anyone selling you the second half as a guarantee is describing a lever they do not have.

If you want a second opinion on where you stand before any of this, [compare your site against a competitor's](https://lantad.co/compare) side by side, or read [the crawlability study](https://lantad.co/research/crawlability-study) for what the failure rate looks like across a population of sites rather than one.

## How to check any of this yourself, for nothing

Every claim on this page about the page layer is testable in a browser in about five minutes, which is deliberate: an argument about evidence that has to be taken on trust is not much of an argument.

[The robots.txt tester](https://lantad.co/tools/robots-txt-tester) shows which of the named AI crawlers your robots file allows, token by token. [What GPTBot sees](https://lantad.co/tools/what-gptbot-sees) fetches your page the way a no-JavaScript crawler does and shows you the text that survived. [The llms.txt checker](https://lantad.co/tools/llms-txt) tells you whether that file exists and parses. [The AI crawler list](https://lantad.co/tools/ai-crawlers) names the bots, who runs them and what each is for.

None of them ask for an email address. If the results are boring, you have your answer and you did not spend anything to get it, which is a better outcome for you than a subscription and a worse one for us. That trade is the point.

Further reading, if the category itself is new: [the glossary](https://lantad.co/glossary) defines the terms both vendors use, [how to get cited](https://lantad.co/how-to-get-cited) covers the per-engine specifics for ChatGPT, Claude, Perplexity and Google AI Overviews, and the other comparisons on this site cover [Profound](https://lantad.co/vs/profound), [Peec AI](https://lantad.co/vs/peec), [Otterly](https://lantad.co/vs/otterly) and [Ahrefs](https://lantad.co/vs/ahrefs).

## Questions and answers

**Is Lantad a Semrush alternative?**

No, and replacing Semrush with Lantad would be a bad trade. Semrush is a broad suite whose AI Visibility Toolkit reports how AI answers describe your brand, alongside keyword, backlink and rank tooling Lantad does not have and is not building. Lantad does one upstream job: measuring whether AI crawlers can reach and read your pages, and naming the fix when they cannot. If you already pay for Semrush, keep paying for Semrush.

**Semrush already has an AI Search Site Audit. Does that not cover crawler readability?**

It covers part of it, and this page says so rather than pretending the feature does not exist. The AI Search Health widget in Site Audit flags technical issues that might prevent AI bots from crawling your content, on up to 100 pages on the Base plan. What it does not do is check robots.txt and the live server response separately per crawler, so a CDN bot rule that contradicts an allowing robots file stays invisible, and it does not fetch the page without JavaScript and diff that against the rendered version, which is the failure that produces a clean audit and an empty crawler view.

**Which tracks more prompts for the money?**

Neither, cleanly, because they buy different things. Semrush's Base plan checks 25 custom prompts every day. Lantad's Starter plan checks 100 distinct prompts once a week and Business checks 250 twice a week. Semrush gives you cadence on a few questions; Lantad gives you coverage of many. Pick by which failure you are worried about: an answer changing, or a question you were never in.

**Can I use both?**

Yes, and that is the recommendation for most teams already inside Semrush. Run the free Lantad scan first to settle whether crawlers receive the page, fix anything it finds, re-scan to confirm, and let Semrush's monitoring report what AI answers do afterwards. Fixing readability first is what gives monitoring something real to measure.

---

Lantad measures whether AI crawlers can actually read a page: it fetches as a non-rendering
crawler, renders as a browser, and reports the gap. Free scan, one URL, no signup.

Method and weights: https://lantad.co/methodology | All pages as markdown: https://lantad.co/md | Crawler policy: https://lantad.co/bot