It stopped calling itself a writing tool, and kept the writing
Frase now describes itself as a content operating system for AI search. The loop is the pitch: research a query against the pages currently ranking, draft against that brief, publish into your CMS, then watch whether ChatGPT and Google AI cite the result and get told when the page starts to decay.
That is a bigger product than the one most reviews describe, and it is the reason it holds the specialist role in the best AI writing tools for blog posts at a score of 7.6 under how we rate. It is also the reason to read the tier table carefully, because the loop is not the same size on every plan.
The entry plan is the whole product at a tenth of the volume
$49 per month, or thirty nine on annual billing, buys one seat, one site, ten articles and fifty audit pages a month. Nothing is feature gated at that price: content scoring, GEO optimisation, the publishing agent and visibility tracking are all there.
Ten articles for forty nine dollars is roughly five dollars an article. Against a general writing tool at seven or ten dollars a month for unlimited output, that looks expensive until you count what you would otherwise buy separately, which is a research tool, a scoring tool and a rank tracker. Against Rytr, Frase costs six times more and does a different job. Against Jasper at per seat pricing, it is competitive and better aimed if search is why you publish.
The eleventh article
Here is the detail that decides which tier you actually need. Professional and Scale sell overages, at five and four dollars per extra article. Starter does not.
Starter stops, it does not meter
On the entry plan the eleventh article of the month is not available at any price. The next step is Professional at one hundred and twenty nine dollars a month, which is a two and a half times increase to solve a one article problem. We compared the published pricing page rather than provisioning an account and hitting the cap ourselves.
If your publishing cadence sits anywhere near ten pieces a month, budget for the middle tier from the start, or accept that some months end early.
Which AI engines you can watch depends on what you pay
Visibility tracking is tiered. Starter watches ChatGPT and Google AI. Professional adds Perplexity. Claude and Gemini arrive on Scale, at two hundred and ninety nine dollars a month.
That is a defensible way to price a monitoring product and an awkward one for a buyer, because the engines you care about are decided by your audience, not by your article volume. A small site whose readers use Claude has to pay Scale money for Starter volume.
Where it publishes, and the one thing the pages disagree on
Publishing goes straight into WordPress, Webflow, Sanity, Wix or FraseCMS, which is bundled with a free hosting tier and paid add ons above it. That is a real advantage over tools that hand you a document and stop.
One inconsistency to check before you commit. The pricing page lists custom API access as an Enterprise feature, while the homepage advertises API and CLI access alongside integrations with other AI tools. If you plan to script anything, ask which is true for your tier before you buy.
Who should buy it, and who should not
Buy it if search is the reason you publish, if you want the brief built from live SERP data rather than from model memory, and if you now also need to know whether AI assistants quote you. The scoring gives a content team something concrete to argue about, which is worth more than it sounds.
Skip it if you publish occasionally, if rankings are not the point, or if you are looking for a cheaper drafting tool. And if what you actually want is the visibility half rather than the writing half, compare it against Writesonic first, because that product sells the same monitoring at a higher entry price. Read how these tools actually work before paying for any of them, and the rest of the category sits under AI tools.
Prices, limits and tier contents on this page were checked against the vendor pricing page and homepage on .
