Jev for SEO and GEO: how to use it, what it costs, how to get access

What is Jev: a 45-second explainer for marketers

ChatGPT answers a question by writing a paragraph. Jev answers by picking from a list. Ask both "is this invoice fraud?" ChatGPT types out a sentence in 8 seconds. Jev gets three options, Fraud / Clean / Review, and returns a score for each in 0.1 seconds: Clean, 88%. No writing, no explaining, just the pick. That is Jev: a model that only decides, 100x faster and far cheaper. Every task below is a pick from a list.

What you can do with Jev: 3 demos

Jev is good at one thing: going through thousands of items and sorting them fast and cheap. Ads, pages, AI answers, it does not matter. You tell it what to look for, it scores every item, and the ones that pass go on to the next step. Three demos below, one per channel, on sample accounts, so the numbers on screen are examples.

Meta Ads: 724 live ads from 37 brands, labelled in 40 seconds

Not our demo: Matthew Berman ran Jev on 724 live ads from 37 brands. Every hook, format, offer, CTA and awareness stage labelled, landing-page mismatches flagged, about 9 cents of tokens.

Meta Ads: find what competitors keep running, then write ads like it

Jev reads 1,600 competitor ads, works out which styles keep running, and scores new briefs in those styles. Whole run: about $0.71.

• Pull 1,600 competitor ads from the Meta Ad Library.
• Tag each ad by type (talking head, myth-buster, price anchor) and check how long it stayed live. Ads that survive 60 days are the ones making money.
• Write and score. An LLM drafts new briefs in the winning styles; Jev scores each on hook, brand fit and how likely it is to survive.

SEO and GEO: audit the whole site in one pass

Four checks that usually take an agency a week, run at once. Each one is a yes/no or a score, so a 500-page site costs cents.

• Titles and headings: keep or change, page by page.
• Competitor pages: which ones are worth copying, which to skip.
• Missing pages: the questions buyers ask ChatGPT that you have no page for.
• Citation odds: how likely each page is to be picked up by AI search.

GEO: check if ChatGPT, Claude and Gemini recommend you, and fix the pages that lose

Jev checks 1,450 buyer questions across three AI engines and tells you where you are cited, who is cited instead, and which pages to fix first.

• Ask ChatGPT, Claude and Gemini about 1,450 questions your buyers ask.
• Check every answer for two things: are you mentioned, and if not, who is.
• Fix. For the pages that lose, Jev scores why (too thin, wrong angle, no data) and an LLM turns that into a rewrite checklist, in order.

9 things Jev can do for your ads and SEO

Most of the grunt work in ads and SEO is the same small question asked thousands of times: is this search term junk? do these two pages overlap? did ChatGPT mention us? Jev answers those questions at scale, for cents.

The rule: if the answer is a pick, a score or a yes/no, Jev can do it. If the answer has to be written out, that is still an LLM job.

What Jev is, in plain terms

Jev is made by TypeSafe AI, a company started by Diogo Almeida, one of the people behind ChatGPT at OpenAI. It opened to early users on September 15, 2026. You give it something to look at (a search term, an ad, a page, an AI answer) and a question with a fixed set of answers. It picks one and tells you how sure it is. That is all it does.

Two things to keep in mind. The accuracy and price numbers come from TypeSafe, not from an independent test. And TypeSafe says it cannot promise the launch price will hold. Plan on today's price, not a permanent one.

The 9 tasks, one by one

1. Search-term triage

What it does: reads every search term in your Google Ads account and sorts it: buyer, researcher, job seeker, competitor, junk. The junk gets added as a negative keyword.

This is the biggest pile in any account, 5,000 to 50,000 terms a month for a mid-size advertiser, and the one most teams check once a quarter if at all. At Jev prices it runs every night.

2. Creative tagging

What it does: looks at each ad, yours and your competitors', and tags the hook, the format, the offer and who it is aimed at. Then you can see which combinations actually win.

Four tags per ad, one pass. This is the task in the Meta Ads demo above.

3. Landing page match

What it does: scores whether the page delivers what the ad or the search result promised. Pages that score low get fixed before you spend more on them.

The one check paid and organic share. The same scorer runs on your ad landing pages and on the pages ranking for your top Search Console queries.

4. Internal link map

What it does: for every pair of pages on your site, answers one question: is there a real reason to link these two? Yes gets a link, no gets skipped.

This is how a 500-page site gets a sensible internal link structure without a person clicking through it. It is the SEO demo above.

5. Cannibalization

What it does: finds pairs of pages competing for the same search, and tells you which pairs to merge.

Search Console shows you which pages share keywords. Jev decides whether they actually answer the same question or just look similar. The merge list also gives your ads one clear landing page per topic.

6. Thin-page gate

What it does: scores each new page 1 to 10 on whether it says anything the rest of your site does not. Low scores stay unpublished until fixed.

The safety check for sites publishing pages at scale. Google's spam updates ask the same question afterwards, so it is cheaper to ask it first.

7. Citation checks

What it does: reads AI answers from ChatGPT, Gemini and Perplexity and checks whether your brand is recommended. You get a citation rate per engine, per topic, per week.

Pulling the answers is easy. Reading thousands of them is what nobody does today. At this price, a 40-question tracker becomes a 4,000-question tracker.

8. Who got cited instead

What it does: when you are not mentioned, names the competitor who was. You get a share-of-answer table by competitor.

Same answers as task 7, second question. Put the table next to your auction insights and you can see whether the competitors outbidding you are also the ones the AI engines recommend.

9. Converting terms with no page

What it does: takes the search terms that already produce sales in Google Ads and checks whether your site has a page for each. The ones with no page become your content calendar.

This is where ads and SEO meet. Paid search already tells you which questions lead to a sale. Each one without a page is a brief for organic and AI search, ranked by the money it has already made.

The pattern behind every use case

Every use case on this page follows one architecture. Code finds the candidates, Jev judges them, a threshold decides what happens, and a frontier model writes only what passed.

This is the same split Ryze AI already uses between changes it makes on its own and changes that wait for approval. A cheap decision model widens the first group without touching the rule for the second.

What it costs at the list price

These are arithmetic at TypeSafe's published price of $0.042 per 1M input tokens with free output. They are not measured runs, and token counts depend on how much context you put in each state.

The point of the table is the order of magnitude. Tasks that cost tens of dollars per run on a frontier model, and therefore run monthly, cost cents and can run nightly.

What Jev cannot do

Jev is narrow by design, and the narrow part matters as much as the price.

• No rationale. You get 0.91, never a sentence explaining it. Anything a client or a finance team will read still needs a frontier model to write the reason.
• Bounded answers only. If the set of possible answers is not known up front, it is the wrong tool. Strategy, copy and diagnosis stay with an LLM.
• Accuracy is triage-grade. 67.8% agreement on the vendor's own benchmark is fine behind a confidence threshold and wrong for auto-applying a budget change.
• "Cannot hallucinate" means it cannot break the schema. It can still pick the wrong option. Measure it against human labels on your own data before trusting a threshold.
• Waitlist and pricing risk. There is no general availability date, and the price may move.

Keep the frontier model for writing ads, pages and answer-first rewrites, for diagnosing a drop once it is routed, and for every explanation a client reads. Keep a human on budgets and bids over 20%, edits to pages that make money, and any reply posted in a community.

How to start before you have access

Jev is in early access. There is no public self-serve signup yet.

More on Jev

• Jev for SEO: 9 site tasks, from internal links to redirect maps.
• Jev for GEO: citation checks across ChatGPT, Gemini and Perplexity at scale.
• Jev for Google Ads: search-term triage and the rest of the ad account.
• Jev for Meta Ads: creative tagging and Ad Library scans.
• Jev for marketing: the full list across ads, SEO, GEO and leads.

For the product side, the Ryze help center covers what generative engine optimization (GEO) is and how Ryze tracks your brand in ChatGPT, Gemini and Perplexity.