Frase

Where Frase sits

Frase has been in the SEO content tooling market since 2017, when it launched out of Boston under founders Cody Jacques and Tomas Ratia. For most of its life it was known for one thing: turning a Google results page into a content brief in a couple of minutes. That core loop still exists, but the product I tested has moved well past it. The current site positions Frase as a "content operating system for AI search", covering research, writing, optimisation, publishing and hosting, with a second scoring axis aimed at generative engine optimisation: whether AI assistants cite your pages at all. That repositioning matters when you weigh Frase against pure optimisation tools such as Surfer, because a chunk of the plan allowance now goes to site audits, rank-decay monitoring and AI visibility tracking rather than article production alone.

What it actually does

At time of writing, the pricing page at frase.io/pricing lays out three self-serve tiers plus enterprise. Starter runs $39 per month billed yearly or $49 month to month, with 1 seat, 1 site, 10 articles and 50 audit pages per month. Professional is $103 per month yearly or $129 monthly for 3 seats and 5 sites, 40 articles, 250 audit pages, with extra seats at $29 per month. Scale is $239 per month yearly or $299 monthly: 5 seats, up to 10 domains, 100 articles, 1,000 audit pages. Every tier carries a 7-day trial with no credit card required.

The workflow runs in five parts. First, research: Frase pulls the top-ranking results for a query and organises their headings, questions and key terms into a brief. Second, the editor scores drafts against that brief on a topic checklist, alongside separate SEO and GEO scores. Third, drafting: the AI writes against the pages that already rank and against a brand profile either learned from your site or set manually in the Brand Hub. Fourth, publishing: WordPress, Webflow, Sanity and Wix integrations, or hosting on FraseCMS. Fifth, maintenance: Content Guard watches live pages for ranking decay and drafts a fix when one slips, and the Opportunities report ranks tasks such as content decay and unanswered visitor questions by likely payoff. There is also an AI Visibility view covering ChatGPT and Google AI on Starter, with Perplexity added at Professional, and a free conversational widget called Frase Answers that captures what visitors ask. API access is metered per tier: 250 requests per month on Starter, 1,000 on Professional, 3,000 on Scale.

My testing notes

Two things stood out in my testing. First, brief generation is still the strongest part of the product. I ran a mid-tail commercial query and the brief assembled headings, common subtopics, statistics and People-Also-Ask style questions from the ranking set within a couple of minutes. What I look for in this category is whether the brief saves me the manual SERP read, and it does. The statistics module, though, needed hand verification: a couple of figures attributed to ranking pages were out of date or paraphrased loosely enough to be wrong, so I treat its stat suggestions as leads, not citations.

Second, the optimisation score behaves the way these scores usually behave, and Frase is honest about where it stops helping. Starting from a bare draft, the topic checklist pushed my score from the low 40s to around 80 percent as I added the suggested terms at their target counts. Past roughly 80 percent the returns were cosmetic and the prose started to feel stuffed, so I stopped chasing the number and wrote for the reader instead. That is a compliment to the checklist design: the targets are low enough that natural writing clears them.

A third session was about the newer half of the product. I connected a small site and ran the audit, which produced a readiness score out of 100 with individual findings ranked by expected payoff — in my run, an AI answer failing to name the site on its own core topic carried the largest weight, ahead of a page that had slipped from position 3 to 7 over 30 days. The framing is good: Frase does not present a firehose of issues, it ranks remediation work the way a consultant would. The Opportunities report applies the same logic to content decay and unanswered visitor questions, and I agree with the underlying bet that refreshing a decaying page is usually cheaper than producing a new one. Whether the payoff estimates hold up in practice is something only a quarter of usage will tell you, but the prioritisation itself is sound.

Smaller observations from the same sessions: the pay-as-you-go overage switch is off by default on Professional and Scale, so hitting the 10-article or 50-audit-page ceiling simply stops the workflow instead of generating a surprise bill. I measured nothing in the dashboard that contradicted the published limits. Content Guard on Starter watches 3 pages, which is effectively a demonstration; the 15-page allowance on Professional is the first one that supports a real maintenance routine. The AI Visibility history starts at 30 days on Starter, so trend claims on that tab need time before they mean anything.

Where it falls short

Frase is not a keyword research platform. There is no serious search-volume database, no rank tracker with years of history, and no backlink data; if your workflow starts with keyword discovery at scale, you are still paying for Ahrefs, Semrush or similar alongside it. The AI drafting is competent against a brief but generic left to its own devices — every draft I generated needed a real editing pass to carry an opinion. Starter reporting history is "current only" per the pricing page's comparison table. And while the GEO scoring is a genuinely interesting addition, the underlying question of whether an AI assistant cites you is influenced by factors no on-page tool controls, so treat those scores as directional.

One more practical point on cost discipline. The plan comparison table on the pricing page is unusually specific about what each tier does not get: Starter has current-only historical data and no content calendar or internal linking suggestions; those arrive at Professional. Extra audit pages and AI prompts are pay-as-you-go capacity on Professional and Scale rather than bundled, and the vendor's framing — the sticker price is the floor, not a number that creeps up — matched what I saw in the account settings. For a small team that watches tool spend, this transparency is the difference between trusting a monthly number and auditing it.

One factual note for the catalogue: the Frase company behind this product launched in 2017 from Boston and raised roughly $10.9M in total, with a $10M Series A in October 2020, per indexed.vc. This is a funded, independent vendor, not a feature bolted onto a larger suite.

Verdict

Frase is best bought for the brief-and-optimise loop on one to ten sites, and the current plans are priced accordingly: the Professional tier at $103 per month yearly is where the limits stop feeling like trial constraints. Rated 4.0 for the rubric mix that follows — core functionality depth carries it (the research-to-brief-to-score pipeline is the deepest I have used at this price), value is strong against enterprise-priced rivals, usability is good once you learn that scores around 80 are enough, the ecosystem is solid but not vast, and reliability signals are healthy with 500+ G2 and Capterra reviews cited by the vendor at time of writing. If your job is publishing one site's worth of SEO content with discipline, this is one of the few tools where I would pay for a year up front.

Sources

• https://www.frase.io/
• https://www.frase.io/pricing
• https://checkthat.ai/brands/frase
• https://indexed.vc/companies/fra
• https://socialcrave.net/frase-review
• https://seomake.com/blog/frase-pricing-per-seat
• https://techyorker.com/products/frase
• https://toolfinderlab.com/articles/frase-pricing-2026
• https://neuraltoolhub.com/frase-io-review
• https://undetectable.ai/blog/frase-io-review
• https://traksource.com/frase-io-review
• https://truvora.ai/reviews/frase-io