Jev for ads
Paid media is mostly judgment at volume: which hook, which angle, which of these 2,000 competitor ads is worth copying. 11 builds here hand that to Jev, and the numbers are startling — a whole brand ad library labelled in 19 seconds for twelve cents, and a 30-persona focus group over 723 ads for 22.
Updated 22 Sept 2026 · by Made with Jev
In short
- 1,891 ads labelled by funnel stage and style: 19 seconds, $0.12.
- 724 live ads broken down by hook, format, offer, awareness stage and landing-page mismatch: about 40 seconds, about $0.09.
- 21,690 scroll-or-stop judgments — 723 ads against 30 personas — for 22 cents.
- Jev never writes the ad. It decides which one, and hands the making to something that generates.
The published runs, side by side
Every row is someone’s own measurement on their own workload, linked to where they published it. They are the reason this page exists: nothing else in the directory clusters this tightly around one job.
| The run | Volume | Time | Cost |
|---|---|---|---|
| A whole brand ad library | 1,891 ads | 19 s | $0.12 |
| 724 competitor ads, broken down | 724 ads from 37 brands | ~40 s | ~$0.09 |
| 700 live ads scanned | 700+ ads | 40 s | $0.04 |
| Every live ad tagged | 430 ads a second | — | $0.60 of spend |
| A 30-persona focus group | 723 ads, 21,690 decisions | — | $0.22 |
| An A/B test against 4,000 personas | 2 posts, 4,000+ profiles | ~12 s | — |
Read the competition's whole ad library
The job that used to be a junior's week: pull every live ad, tag each one by hook, angle, offer, format and funnel stage. The published runs land between 19 and 40 seconds and between four and twelve cents.
Matthew Berman
@TheMattBerman
jev is INSANE. in 40 seconds it broke down 724 live ads from 37 brands. every hook. every format. offer. cta. awareness stage. landing page mismatch. used 9 cents of tokens. (will be avail in @stealads + mcp)
ares. 🎧
@aresotik
ESTA HERRAMIENTA ACABA DE ROMPER TODO EL MERCADO DEL AD SPY Maxfusion ha cogido JEV, el modelo nuevo de TypeSafe, y le ha metido la ad library entera de una marca → 1.891 anuncios clasificados → 19 segundos → 0,12 $ Y no es un resumen: cada anuncio etiquetado por etapa del funnel y estilo creativo, más la radiografía completa de la cuenta Llega pronto al MCP de maxfusion
Ackerman
@Yarilo7brigada
jev is an insanely cool product Saw a post about jev and decided to check it out This thing is straight up gold whoever built it is a genius jev is the future Scanned over 700 live ads in 40seconds flat, and it cost me just 4 cents. If you're in marketing, this kind of crazy fast data crunching is hands down the best thing out there. Doing this through Opus 5 would run around 2 to 4.5 million tokens, set you back $15 to $50, and take anywhere from 10 minutes to an hour. That’s 200 to 500 times more expensive than Jev, and way slower.
zre
@zrebroia
Jev lleva toda tu operación de marketing de anuncios por menos de 3$. → Escanea cada anuncio activo de la competencia y lo etiqueta por hook, formato y tasa de supervivencia → 430 anuncios leídos por segundo. 0,6$ gastados.
佐藤 裕二／さいたまのWEB広告運用者
@Ysato_su04
これができたら衝撃的。 JevがMeta広告ライブラリから長く回っている勝ちパターンを自動抽出、Google検索語句まで分析して不要クエリ除外、LPとのズレまでスコアリングしてくれる。調査から改善まで一気通貫で、人が何時間もかけていた工程を3ドル未満で済ませるのは衝撃的。 広告運用者として実務で感じるのは、クリエイティブ疲れやオーディエンス重複の見落としがROAS低下の主因になること。 このツールなら仮説立案が高速化され、PDCAが30倍速になる可能性大。
XContent and growth
Winning patterns out of the Meta Ad Library
keno
@kenonews
JEV makes competitor research feel like a cheat code. Give it your competitors’ ads. Break them down by hook, angle, offer and format. Then turn recurring combinations into a shortlist for your next creative test. From an endless swipe file to “here’s what we should try next.” Your competitors just became your creative department.
XContent and growth
A shortlist from a competitor swipe file
Test the creative before you pay for the impression
Synthetic pretesting is where the cost per decision changes what is possible: one focus group here is 723 ads against 30 personas, 21,690 separate judgments, for 22 cents.
Matthew Berman
@TheMattBerman
jev KILLED the focus group. it scrolled 723 ads as 30 buyer personalities 21,690 stop or scroll decisions. 22 cents. (will be avail in @StealAds + mcp)
Sim Audience
@SimAudience
Jev is WILD I gave it two launch tweets and fed it over 4000 demographic profiles of real survey participants Twelve seconds later, a simulated A/B test tied to actual personas voting on the best tweet you can just do things i made it 100% free (link below)
SCRAPS
@scrapsonsolana
For marketing $SCRAPS, I’m trying something I haven’t seen anyone else do. I’ve been experimenting heavily with the latest AI tools (JEV from @typesafeai), burning through 178M+ tokens and nearly 48,000 requests testing different ideas. One of those experiments is focused entirely on marketing: finding the fastest path to high engagement and the best ROI possible. Now that SCRAPS is live on Google Play, it’s time to put it to the test. The experiment starts now.
Pick and filter the creative itself
Jev does not make the ad. It decides which asset goes in, which cut survives and which draft is worth producing, and hands the making to a model that generates.
Higgsfield AI 🧩
@higgsfield_ai
Jev is really good at content filtering and asset selection. DeepSeek + Higgsfield turn the selected assets into ad creatives.
XContent and growth
Ad creatives from filtered assets
内田勉 DirecTune.app β公開中
@sidodtv
なんか24時間後に Tiboリセットくるらしいので、今のうちに #directune の開発を進めるよ〜！ まずは、ゆっくり風動画解説で背景が自動作成されるようになったよ。試しに Jevの解説動画を作ったよ DrecTuneは、AIがアシストしてくれる動画生成・制作サービスだぞ‼️
XContent and growth
Auto-generated backgrounds for a Jev explainer
How to build the teardown yourself
The shape is the same in every one of these builds. Pull the ads from wherever they live — a Meta Ad Library scrape, a swipe file, a screenshot run through OCR. Send one ad as the state. Ask every question you care about in the same call: which hook type, which awareness stage, which offer, does the landing page match, is this the kind of ad that has been running for months. Then branch on the confidence, keeping the sure labels and sending the rest to a person or to a larger model.
The mistake that costs money is one call per question. Questions in a call are answered in parallel against the same ad, so a twelve-point teardown costs roughly what a one-point one does. The request itself is on how to use Jev.
What to be careful about
Synthetic personas are a prior, not an audience. A model deciding scroll or stop for an invented profile tells you something about the creative and nothing about your market; the builds that do this say so themselves, and the honest use is ranking a batch before you spend, not replacing the test. Ad copy is also short, which means your bill is dominated by how much surrounding context you send, so trim the state before you trim the rubric.
Next door: Jev for SEO for the organic side, Jev for marketing for the rest of the function, and the pricing page for where all these figures come from.
Common questions
- What can Jev do for advertising?
- Read and label creative at a speed that changes the workflow. Competitor ad libraries tagged by hook, angle, offer, format and funnel stage; synthetic personas voting on a draft before it runs; assets filtered before production. It does not write the ad or make the image.
- How fast can it read a competitor's ad library?
- The published runs: 1,891 ads labelled in 19 seconds for $0.12, 724 ads broken down in about 40 seconds for about $0.09, more than 700 live ads scanned for four cents, and one build reporting 430 ads a second.
- Can Jev pretest creative?
- Several people do exactly that. One build runs 723 ads past 30 personas as a scroll-or-stop decision — 21,690 separate judgments — for 22 cents. Another put two launch tweets to more than 4,000 survey profiles in about twelve seconds. Treat the output as a prior, not as an audience.
- Does Jev make ad creative?
- No. It picks and filters. In the builds here it chooses which assets go forward and which drafts are worth producing, and a generative model or an editor makes the thing itself.
- What does an ad teardown cost with Jev?
- Between four and twelve cents for the published runs of 700 to 1,900 ads, at the list price of $0.042 per million input tokens with output free. The variable is the length of the ad copy you send, not the number of questions you ask about it.
Made with Jev is independent and not affiliated with TypeSafe AI. Every figure on this page is the one its author published, linked to where it can be checked.