Frase Review November 2026: From an AEO Consultant

Frase is the better fit when one subscription has to research a page, draft it, score it, and publish it to WordPress, Webflow, Sanity, Wix, or FraseCMS, and a small daily sample of AI answers is enough. It is a weak first buy when the question the board asked lives in Copilot, Grok, or DeepSeek, because those three names sit on the Enterprise row, or when someone reads the feature page's engine list as if Starter already included Perplexity, Claude, and Gemini. On 15 November 2026 the pricing page showed Starter at $39 a month billed yearly, Professional at $103, and Scale at $239. The monthly rates printed beside them were $49, $129, and $299. Enterprise was custom.

Last updated - November 2026

I am Austin Heaton, an AEO consultant. The content programs I am asked to look at are mostly fintech. The rest are B2B SaaS and a few other industries when the calendar is the same problem. A writing login is useful when I can see the question, the cited URL, and the sentence a person still has to approve. A GEO score with no stored answer is not a plan, and a visibility chart that cannot name the page is not a brief.

Prices, the comparison table, the overage rates, the trial caps, and the FraseCMS add-on are from Frase's pricing page, read on 15 November 2026. The engine list, the buying-stage definitions, the citation fields, and the statement that AI Visibility monitors and does not edit the page are from Frase's AI Visibility page. How a check is counted, how locations spend prompt slots, what the four overview figures mean, and which history windows are locked are from Frase's AI Visibility and GEO documentation. The same plan dollars, described as 20% off the monthly rate when billed yearly, are repeated on Frase's product reference. Where a cell below is Frase's marketing sentence rather than a cap I could count, I say so. I did not create an account. I did not run a prompt. I did not publish a page.

The side-by-sides already on this site are not being rewritten as a second invoice. Frase beside Surfer, from a 3 November read, is Frase vs Surfer for blog posts. I am citing the 15 November pricing page for the dollars below, not that older table. Surfer on its own is the SurferSEO review. Frase beside MarketMuse is MarketMuse vs Frase. A grader that stops at the draft, rather than a system that also publishes, is the Clearscope review. The measurement program around any of these logins is the work behind measuring AI visibility.

Key Takeaways

• On 15 November 2026 Frase's pricing page showed Starter at $39 a month billed yearly ($468 a year), or $49 month to month. Professional was $103 billed yearly ($1,236 a year), or $129 month to month. Scale was $239 billed yearly ($2,868 a year), or $299 month to month. The FAQ says yearly billing takes 20% off the monthly rate. Enterprise is custom. Every self-serve plan offered a 7-day trial with no credit card.
• Articles are 10, 40, and 100 a month. Audit pages are 50, 250, and 1,000. AI generations, which the table describes as on-demand actions such as idea expansions and image generation, separate from the article count, are 25, 100, and 350. Starter holds at those limits. Professional and Scale can turn on pay-as-you-go, off by default. The FAQ prices that overage at $5, $0.50, and $0.25 on Professional, and $4, $0.40, and $0.20 on Scale, per extra article, audit page, and what the FAQ calls an AI prompt.
• AI Visibility prompts are 20 on Starter, 50 on Professional, and 75 on Scale. Starter tracks ChatGPT and Google AI. Professional adds Perplexity. Scale adds Claude and Gemini. Enterprise adds Grok, Copilot, and DeepSeek to that set. The AI Visibility page lists all eight engines and says how many you track depends on the plan. I am using the pricing table for what a card includes.
• Publishing to WordPress, Webflow, Sanity, Wix, or FraseCMS is marked on every plan. Content Guard, which watches Google rank and traffic decay and proposes a fix, covers 10, 75, and 300 pages. Approval is the default. The agent can publish on its own only when you switch that on. AI Visibility alerts you. The pricing FAQ says it never touches the content.
• Seats are 1, 3, and 5, with extra seats at $29 a month on Professional and Scale. Domains are 1, 5, and 10. Client-ready exportable reports start on Scale. White-label reports, a branded client portal, SSO and SAML, and a dedicated manager with an SLA are Enterprise. Historical data on the table is current only, 3 months, 12 months, and 24 months, with a note that AI Visibility keeps its own history, from 30 days on Starter. No plan on this page guarantees a citation.

What Is Frase in November 2026?

Frase sells one loop and then meters how much of it you can run. The product reference names the loop Listen, Create, Publish, Optimize, Fix. Listen is a conversational widget, Frase Answers, embedded on your site. Create is the agent that researches, drafts, and scores. Publish is the CMS handoff, or Frase's own host. Optimize is an SEO score and a GEO score on the draft, plus the AI Visibility log. Fix is Content Guard, which watches a published URL for Google decay. The pricing page is narrower than that story, and the pricing page is the invoice.

That loop is the reason Frase is not a pure visibility log. A tracker that only stores answers is a different category of product. Frase's AI Visibility page says the tracking and the content production sit in one subscription, and it walks a gap through five steps: research the question, write the draft, check a GEO score beside an SEO score, wait for your approval, then publish. I did not watch those five steps run. I am citing the page. A generated recommendation and a URL on your domain are still different objects until the approval happens and the CMS accepts the post.

Two sentences on the pricing page should stay next to each other. The create block says nothing ships until you approve it. The plan cards say the agent publishes on its own when you let it, on any content type. The FAQ explains the switch. During the trial you can turn autonomous publishing on for a content type, watch what it ships, and turn it off. Content Guard is a second switch. Approval is the default on every Content Guard change, and you can opt in to let Frase ship low-risk fixes. AI Visibility is not either switch. The FAQ says it monitors AI engines and alerts you, and it never edits the content. I would not put Frase in a statement of work as a system that publishes unattended unless the order records which switch was turned on, and for which content type.

The pricing page also prints a 4.8/5 rating on G2, and it prints customer quotes. One quote praises emails about AI visibility. Others praise research speed, the audit, and the MCP. I am not using those sentences as results I measured. A quote on a pricing page is a customer saying something. It is not a before-and-after I recomputed. The product reference says thousands of teams use Frase. I have not audited the count.

The job I can defend is narrower than the hero. Someone has to choose the questions a buyer actually asks, someone has to stop a wrong fee, and someone has to know whether the cited URL was yours. Frase's table can hold a small prompt set, store a cited URL, and draft a page into a CMS you already use. It does not, on the pages I read, close the fee check by itself. Sentiment is a figure in the documentation. It is not a compliance review. The tool can sample. It does not know which prompt the pipeline needs a model to answer.

How Do the Frase Plans Compare?

The table is the pricing page as I could read it on 15 November 2026. "Google AI" is Frase's label. I am not expanding it into Google AI Overviews plus Google AI Mode, because the pricing table does not. The product reference, in its opening description, says Frase aims to get cited by engines "like ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews." That sentence is a category description. The plan columns are the invoice. If a demo shows "Google AI" pointing at one Google surface and not the other, the demo wins. I did not see that demo.

If the 20 prompts and the two engines are not the questions the sales team hears, Book a Call and bring one prompt. The prompt shows which plan is even eligible before anyone argues about Scale.

The article cap and the prompt cap are different meters, and a team that buys Frase for the blog will feel the first one before the second. Starter at 10 articles is a thin calendar. A fintech site that publishes a comparison, a glossary page, and a product update in the same month is already at the edge of that card, before anyone refreshes an old URL. Professional at 40 is the first plan where a small in-house team can hold a real monthly quota and still have a content calendar, which the table withholds from Starter. Scale at 100 is the agency volume, across up to 10 domains, not a way to buy Grok. Grok is Enterprise. Paying for Scale does not put Copilot on the engine list.

AI generations are a third meter. The table says they are on-demand actions the agent runs, such as idea expansions and image generation, and that they are separate from the monthly article count. Starter's 25 is easy to spend if every article asks for a hero image and a round of expansions. I would not assume a generation and an article are the same object, and I would not assume a generation and an AI Visibility prompt are the same object either. The overage FAQ uses the words "AI prompt" for the third rate. The table uses "AI generations" for the included count. I am not merging those labels into one forecast. Ask the billing screen which meter the $0.25 hits.

Which Engines the Prompt Cap Can Actually See

The engine list is the decision, and it is a ladder, not a menu you pick from. Starter runs ChatGPT and Google AI, and it runs 20 prompts. Professional adds Perplexity and raises the cap to 50. Scale adds Claude and Gemini and raises the cap to 75. Enterprise is the column that adds Grok, Copilot, and DeepSeek, with a custom prompt count. There is no add-on card, on the page I read, that sells one extra engine onto Starter the way some visibility tools sell an extra model. If Perplexity is the engine the weekly report already uses, Starter is the wrong card even when 20 prompts would have been enough.

The AI Visibility page is wider than any one column. Its capability table lists ChatGPT, Google AI, Perplexity, Claude, Gemini, Grok, Copilot, and DeepSeek, and the FAQ on that page says Enterprise covers all eight. I counted eight names. I am not going to brief a founder that a Professional trial, or a Professional subscription, tracked Claude because the feature page named Claude. The trial FAQ is explicit in the other direction. The trial includes AI Visibility on ChatGPT and Google AI, even though the trial is described as Professional's feature set. Perplexity is a Professional inclusion on the paid table. It is not in the trial sentence I could read. I would confirm the engine toggle inside the trial before I reported a Perplexity gap from a seven-day sample.

Google AI is the label I will not inflate. The pricing table says Google AI. The documentation, in the section on Search Console, says AI Search uses your Search Console data to show which of your pages and queries Google's AI features find when they search Google. That is a Search Console report. It is not the same object as a prompt Frase sends to an engine. I will not tell a client that a Starter plan tracked Google AI Mode and Google AI Overviews as two rows. I did not find those two names as separate engines on the pricing table. If the board's sentence is "AI Mode," the honest status is that Frase's invoice says Google AI, and the demo has to show which surface that label hits.

Locations spend the prompt cap. The documentation says each prompt counts once per location toward the plan's prompt limit, so one prompt asked in three locations uses three prompts. Starter's 20 can be ten questions in two countries, or four questions in five countries, or twenty questions in one country. It cannot be all three of those shapes. The AI Visibility page says per-prompt country targeting covers 40 or more countries, and that answer retrieval reflects the target market. Capable and included are different rows. The country list is a targeting option. The prompt cap is the budget. A US product and a UK product and a German product will burn Starter before the category questions are finished, if each question is asked in each country.

The documentation also describes a places workflow for businesses with several branches or city pages. Frase lists places it found on the site, ordered by the search traffic each page gets, and you can track up to as many as you have open prompt slots. Each place's question is asked with its region, so two cities with the same name are not mixed. I did not run that workflow. I am citing the doc. A local brand with a long city list does not get those cities for free. They get as many as the prompt cap has left.

History is a second gate on the same prompts. The pricing table says historical data is current only on Starter, 3 months on Professional, 12 months on Scale, and 24 months on Enterprise, and the caption says AI Visibility keeps its own history, from 30 days on Starter. The documentation says the date picker offers 7, 30, 90, 180, or 365 days, or a custom range, and then it warns that the last 180 days requires Scale or Enterprise, and the last 365 days is Enterprise only. Scale's table cell says 12 months. The doc says 365 days is Enterprise only. Both sentences were in Frase's materials on 15 November. I am not deleting the doc to make the table true, and I am not deleting the table to make the doc true. A year of charts is a question for the screen after you are logged in. Scheduled reports, the doc says, always cover the last 30 days. Export and scheduled reports in AI Visibility are locked on Starter. A Starter login can look at the current window. It is not the plan that emails a PDF.

What Frase Counts as Visibility, a Mention, and a Citation

The documentation is the definition I would put on a slide, and it is a definition of the prompts you are tracking, not of a market Frase invented. Once prompts are set up, the Overview tab leads with four figures. Visibility is the share of your measured answers that named or cited you. The doc says each check of a prompt on one platform is one answer. Share of voice is, of every brand named in your answers, how often it was you. Average position is your rank on visibility and on share of voice among the brands you track, averaged. The doc says 1.0 means first on both. Sentiment is how positively AI answers describe you. Clicking it shows the words answers use and quotes from the answers. I did not click it. I am citing the definition, not a screenshot.

That visibility definition mixes two events. An answer that names the brand and an answer that cites the brand both count toward the share. The AI Visibility page is more careful about the objects underneath the score. It says a brand mention tells you that you are present, and a cited URL tells you which page earned a place in the answer. Frase tracks the page and its position in the citation list, and the other sites the answers lean on, from review sites to forums. Those are the fields I would actually open. A share-of-voice movement without the URL is how a wrong fee gets into a slide. The page also says industry benchmarks compare your citation rate with the industry median, per engine, week by week, with a quartile placement and an industry share of voice for domains you did not add as competitors. I did not see the industry list. I would not quote a quartile in a board deck until the screen shows which set of domains produced it.

Buying stage is the sort Frase adds on top of the prompt. The feature page and the doc agree that every tracked prompt is classified as awareness, consideration, or decision. Awareness is the broad category question. Consideration is the problem question. Decision is the comparison and the alternatives question. Frase assigns the stage, and the doc says you can check it in a dropdown before you save a prompt you wrote. Per-stage scores and a stage-coverage view are the report. I would still read the label. A prompt the model files under awareness can be the question a buyer asks the week they buy. The stage is Frase's classification. It is not a CRM stage.

The doc's tabs are the operating picture, and they are more specific than the pricing captions. Overview is the four figures, competitors, platform coverage, the buyer journey, citation sources, and results by location. Trends is the movement. Prompts is every tracked prompt, and opening one shows each platform's answer, the searches the AI ran, the competitors in those answers, and recent answers. Recommendations is what to fix, with the prompts, competitors, and cited sites as the evidence, and a button that starts the work. Alerts list prompts that gained or lost a mention or a citation, minutes after a check finishes, plus share-of-voice shifts, drops, and crawler spikes. Daily tracking with alerts is also on the feature page. I did not receive an alert. The word "minutes" is Frase's claim about latency. I would not put it in a service level.

Crawler logs and AI referrals are the on-site half, and they are not the sentence in the answer. The doc says crawler logs need a tracking snippet on the site, and that AI referrals come from a connected Google Analytics 4 property. The pricing table marks AI crawler monitoring on Professional and above, and not on Starter. I did not install a snippet, and I did not connect Analytics. A log of hits is worth having. It does not tell you whether the model named the brand. Search Console's AI Search view, in the same doc, is a third object: queries Google's AI features already associate with your pages. That is useful when the site already has impressions. It is not a substitute for a prompt you typed.

Competitor benchmarks, on the feature page, are share of voice against competitors you choose, plus mention benchmarking. I did not find a numeric cap on how many competitors a plan allows. I will not invent five, or nine, or unlimited. If the category routinely names more than a handful of companies, the screen has to show whether the share-of-voice math is the share against the logos you added. The industry share-of-voice row is the wider set, and it is a weekly benchmark, not the daily prompt.

Gap analysis is where off-site pages show up. The feature page says Frase shows the other sites the answers lean on, so you know which page to strengthen and where a mention would count. The tracker can name the URL. It does not place a new one. When the missing source is a page on someone else's site, Offsite Posts is a service for getting brand mentions and posts published on third-party sites.

GEO Scores, Content Guard, and the Page That Ships

GEO, in the documentation, is the optimization half, and AI Visibility is the tracking half. The doc says they are related and distinct. The GEO score sits in the editor beside an EEAT score and an SEO score. GEO evaluates four things: whether AI can pull a clear answer, whether the information is specific enough to quote, whether facts and definitions are easy to parse, and whether the content explicitly answers the questions AI systems are asked. The suggestions named in the doc are the usual ones: a definition early, a concrete data point with a source, the answer stated before the explanation, lists and tables, and links to sources you trust. I did not open a scored draft on 15 November. A score moving from one number to another, inside Frase, is not a citation in ChatGPT. The pricing FAQ says the same thing in plainer language. Frase moves the in-app SEO, GEO, and AI-visibility scores as you apply fixes. Rankings and AI citations build over the weeks after you publish. Frase does not promise a ranking or a citation inside a 7-day window.

Content Guard is the refresh product, and it watches Google, not the AI answer. The pricing table says it watches Google rank and traffic decay, proposes the fix, and republishes on your approval. The caps are 10, 75, and 300 pages. A site with a large documentation set will spend Starter's 10 pages on the URLs that already make money, and it will not be watching the rest. The feature page says you choose automatic application only when you want to, and that backlink gaps can come into the same plan. The product reference describes backlink research as a scan of domains that link to competitors and not to you, with an outreach email drafted for you to send. You send the email. I am citing that sentence because a drafted email is not a placed mention. I did not send one.

Publishing is the row that separates Frase from a grader. The pricing page lists WordPress, Webflow, Sanity, Wix, and FraseCMS on every plan, including Starter. FraseCMS has its own prices, and they are add-ons, not a reason to upgrade the content plan. The free tier is included: 100,000 page views a month, 25 GB of bandwidth, structured data and an llms.txt file on every page, and a Powered by Frase badge. Basic is $19 a month, or $190 a year, for 250,000 page views, 100 GB, a path on your own domain, and no badge. Pro is $99 a month, or $990 a year, for 1,000,000 page views and 500 GB. The FAQ says past the cap the free tier pauses for the month, and the paid tiers keep serving at $1.00 and $0.50 per extra 1,000 page views, with a spend cap set at the base price by default. A page view counts when a page is served to a reader or pulled through the API. I would not describe FraseCMS as free at any traffic level. It is free inside the cap, on a Frase host, with a badge, unless you buy Basic or Pro.

The table also says Starter covers the core content formats, and Professional and up unlock the full set. I did not find the format list on the pricing page. I will not guess which format is missing. If the reason to leave Starter is a format, the trial screen has to show the locked type. Omni-channel atomization, turning one piece into posts sized for other channels, starts on Professional. I did not see a channel list with allowances. The product reference names LinkedIn, X, newsletters, and Instagram as the derivatives. I am citing that list as the reference's list, not as a quota I counted.

API, MCP, and CLI requests are 5,000, 20,000, and 50,000 a month, then custom. The product reference points at an MCP server and an API. The pricing page says Frase runs where you already work, through MCP and CLI, on every plan. A request cap is not a feature list. I did not call the API. I would not promise a client that an agent inside another product can publish to their CMS until the request that does the publish is one of the calls the cap covers, and until the approval switch is the one they chose.

Diagrams and infographics are withheld from Starter, then 20 on Professional and 50 on Scale. That is a clean gate. It is also easy to ignore until a draft needs a figure. I am mentioning it because the empty cell is a limit, not a forgotten row.

Who Frase Fits, and Who Should Skip It

Pick from the engine, the article quota, and whether a person will read the draft.

• One site, ChatGPT and Google AI are the engines you report, and 10 articles a month can hold the calendar. Starter, if 20 prompts in one country are enough and you can live without Perplexity, a content calendar, and crawler monitoring. The plan holds at its caps. That is the point of the card.
• You need Perplexity, a second site, or a calendar. Professional. It is also the first plan with internal linking suggestions, AI crawler monitoring, and the atomization row. It is not the plan that adds Claude or Gemini.
• An agency with several domains and a client who wants a file. Scale, if 10 domains, 100 articles, and exportable reports are enough. White-label reports and a branded portal are the Enterprise line, and the pricing FAQ says that line is the reason to talk to sales. Scale does not include SSO.
• You need Copilot, Grok, or DeepSeek on a prompt you wrote. Those names are on the Enterprise cell. I would not buy Scale to get an engine the Scale column does not list. Read the Frase vs Surfer comparison before you treat this ladder as the only way to see a draft and a tracker together.
• You wanted a large daily prompt log and you do not need Frase to write the page. Twenty, fifty, and seventy-five prompts are samples, not a category, once locations multiply. A dedicated visibility log is a different subscription. This page is the content system that includes a sample.
• You need the model fact-checked against a fee schedule. Sentiment and a cited URL are the features I could read. They are not a check that the fee in the answer matches the fee on the site. On a live account, Austin Heaton still reads that sentence before it goes in a deck.
• You wanted the agent to publish the blog with nobody in the loop. The default on the pages I read is approval. Autonomous publishing is a switch, on a content type, inside the article cap. Content Guard's low-risk switch is a second opt-in, and it watches Google decay, not the AI answer. Leave the approval on for any page that states a price, a license, or a partner.

The consultant and agency version of this decision, if the login is not the thing you want to operate, sits on the comparison index. A tool subscription and an engagement are different bills. This page is only the tool.

What I Could Not Confirm

• A stored answer from inside the product. I read the public pages and the documentation. I did not create an account.
• Whether "Google AI" on the pricing table is Google AI Overviews, Google AI Mode, or both. The table does not spell those two names. The product reference uses "Google AI Overviews" in a descriptive sentence, not in a plan column.
• Whether the trial's AI Visibility toggle includes Perplexity. The paid Professional column includes it. The trial FAQ names ChatGPT and Google AI.
• Which ChatGPT model is queried, and whether the capture is the consumer interface or an API. The pages I opened did not say.
• That the overage "AI prompt" in the FAQ is the same meter as "AI generations" in the table, or the same meter as an AI Visibility prompt. Three labels, and I will not collapse them.
• A dollar for the Frase Answers conversation add-on. The table says every plan includes a free tier, one widget per site, and that higher volume is a paid add-on. I did not find that add-on's price. FraseCMS is a different add-on, and its prices are above.
• The content formats Starter withholds. The table says Professional unlocks the full set. It does not name the missing formats.
• A competitor cap, a numeric sentiment scale, and a count of searches inside one answer. The Prompts tab says you can see the searches the AI ran. It does not say how many.
• That Scale's 12-month history cell and the doc's "365 days is Enterprise only" warning describe the same control. I am leaving both sentences in place.
• The 4.8/5 rating and the customer quotes. They are on the pricing page. I did not recompute them, and I am not using them as outcomes.

Frequently Asked Questions

What does Frase Starter include?

On the pricing page I opened on 15 November 2026, Starter is the cheapest published plan. The yearly rate and the month-to-month rate beside it are in the pricing table in this article. It includes one seat, one site, 10 articles, 50 audit pages, and 25 AI generations. AI Visibility is ChatGPT and Google AI, with 20 prompts. Content Guard watches 10 pages. Publishing to WordPress, Webflow, Sanity, Wix, or FraseCMS is included. The content calendar, internal linking, AI crawler monitoring, Perplexity, and exportable client reports are not on that column. The plan holds at its limits instead of billing an overage.

Which AI engines can I track on a self-serve Frase plan?

Starter tracks ChatGPT and Google AI. Professional adds Perplexity. Scale adds Claude and Gemini. The prompt caps on those three columns are 20, 50, and 75. Grok, Copilot, and DeepSeek are on the Enterprise cell, and the AI Visibility page says Enterprise covers all eight engines named on that page. I did not find an add-on that moves one of those three engines onto a self-serve plan. The trial FAQ names AI Visibility on ChatGPT and Google AI, so I would not brief a Perplexity result from the trial until the toggle shows it.

How does Frase count prompts, mentions, and citations?

The documentation says each check of a prompt on one platform is one answer, and that one prompt asked in three locations uses three of the plan's prompts. Visibility is the share of those answers that named or cited you. Share of voice is how often the brand was the one named, among brands in the answers. A cited URL is tracked at page level, including its position in the citation list. I did not open a stored answer, so I have not checked those definitions against a live row. Average position, in the same doc, is the rank on visibility and on share of voice, averaged.

Will Frase publish to my CMS without a person reading the draft?

Only if you turn that on. The pricing page lists WordPress, Webflow, Sanity, Wix, and FraseCMS on every plan. It also says nothing ships until you approve it, and that the agent publishes on its own when you let it. Content Guard waits for approval unless you opt in to low-risk fixes, and it watches Google rank and traffic decay. AI Visibility alerts you and, on the FAQ I read, does not edit the page. Autonomous publishing still spends the monthly article allowance.

What happens when a Frase plan hits its monthly limit?

Starter holds at the limit until the cycle resets or you upgrade. Professional and Scale can turn on pay-as-you-go, and the pricing page says that switch is off by default. The overage rates for an extra article, an audit page, and an AI prompt are in the pricing table in this article, and each rate is lower on Scale than on Professional. The same table calls the included third meter AI generations. I would confirm on the billing screen which meter the overage uses before I forecast it. The documentation says a move to a plan with fewer prompts is blocked until you pause enough tracked prompts to fit.

What is in the Frase trial, and does it need a card?

The pricing FAQ says the trial is 7 days, with no credit card, and that it uses Professional's feature set at a cap of 5 articles, 50 audit pages, 25 AI prompts, and up to 3 seats. AI Visibility in that sentence is ChatGPT and Google AI. You can turn autonomous publishing on during the trial and turn it off again. I did not start the trial, so I am citing the FAQ rather than a workspace I used. Enterprise is a sales conversation, not that trial.

Does a GEO score mean the page will be cited?

No. The documentation describes the GEO score as a read of how extractable, specific, structured, and question-shaped the draft is, shown beside an SEO score and an EEAT score. The pricing FAQ says Frase moves those in-app scores as you apply fixes, and that rankings and AI citations are slower than the score. Frase does not promise a ranking or a citation inside the 7-day trial. A cited URL in the visibility report is a different object from a score on a draft.

Who should skip Frase?

Skip Starter when the engine you need is Perplexity, Claude, or Gemini, because those names start on Professional or Scale. Skip every self-serve plan when the engine is Copilot, Grok, or DeepSeek, because I found those names on Enterprise. Skip Frase when the job is a large prompt log and nobody needs the article quota, because 75 prompts on Scale is still a sample once each location spends a slot. Skip unattended publishing on any page that states a fee. No plan on the page I read guarantees a citation.

Read next:

• Frase vs Surfer for Blog Posts
• SurferSEO Review, November 2026
• MarketMuse vs Frase
• Clearscope Review, November 2026

If the engine ladder, the article cap, and the approval switch are three different purchases, Book a Call and we will put the prompt, the CMS, and the person who reads the fee on one page before anyone upgrades for an engine the card does not add.