Jev for paid ads: where a $0.0004 decision model fits
Jev is TypeSafe's new decision model. It cannot write an ad. It can decide, about 100 times faster and far cheaper than a frontier model, whether a search term is a buyer, whether a creative is fatigued, or whether a change is safe to apply. Here are 8 cross-channel jobs that fit and the ones that do not. Ryze AI publishes this blog and does not run Jev in production yet.
Built by our community of 2,000 marketers
Free skills and prompts for paid ads and SEO
Templates for Claude, ChatGPT and Perplexity.
Used by 2,000+ marketers
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 job on this page is a pick from a list.
Jev on 724 live ads in 40 seconds
Matthew Berman ran Jev on 724 live ads from 37 brands. In 40 seconds it labelled every hook, format, offer, CTA and awareness stage, and flagged the landing pages that did not match the ad, for about 9 cents of tokens. This is the job most ad teams do by hand once a quarter, if at all.
What Jev looks like on an ad account: demo
One recording on a sample account. Jev pulls competitor ads, scores every one, and the winners go to an LLM to write briefs. The counters on screen are examples, not measurements.
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.
Ryze AI — Autonomous Marketing
Try Jev-powered AI Marketer
- Jev makes the small decisions, Ryze AI makes the change
- Connect Google Ads, Meta, GA4 and Search Console
- Free to connect, flat fee for autopilot
2,000+
Marketers
$500M+
Ad spend
23
Countries
The 8 jobs at a glance
An ad account produces thousands of small decisions a day and a handful of big ones. Teams spend frontier-model money, or human hours, on both. A decision model takes the small ones, which leaves the expensive model and the people for strategy, copy and budgets.
- 1Search-term triage (Google)Choice
Question Jev answers: Buyer, researcher, job seeker, competitor or junk?
What happens next: Negatives above the threshold, the rest queued
- 2Creative tagging (Meta, TikTok)Choice ×4
Question Jev answers: Hook, format, offer, awareness stage?
What happens next: Performance grouped by tag
- 3Anomaly triageChoice
Question Jev answers: Is this CPA jump tracking, auction, creative or budget?
What happens next: Routed to the right check before anyone is paged
- 4Change safety gateNoul
Question Jev answers: Is this proposed change reversible and under the account's limits?
What happens next: Yes is applied, no waits for approval
- 5Tracking sanityNoul
Question Jev answers: Do these two conversion counts describe the same events?
What happens next: Gaps over 10% open a tracking row and pause optimization
- 6Lead quality scoringScore
Question Jev answers: Score this form fill 0 to 100 against the ICP
What happens next: Scores are sent back as offline conversion values
- 7Landing page matchScore
Question Jev answers: Does the page deliver what the ad promised?
What happens next: Low scores are fixed before spend
- 8Report annotation routingNoul
Question Jev answers: Does this weekly change need a client explanation?
What happens next: Only the yes rows get an LLM-written note
| # | Job | Answer type | Question Jev answers | What happens next |
|---|---|---|---|---|
| 1 | Search-term triage (Google) | Choice | Buyer, researcher, job seeker, competitor or junk? | Negatives above the threshold, the rest queued |
| 2 | Creative tagging (Meta, TikTok) | Choice ×4 | Hook, format, offer, awareness stage? | Performance grouped by tag |
| 3 | Anomaly triage | Choice | Is this CPA jump tracking, auction, creative or budget? | Routed to the right check before anyone is paged |
| 4 | Change safety gate | Noul | Is this proposed change reversible and under the account's limits? | Yes is applied, no waits for approval |
| 5 | Tracking sanity | Noul | Do these two conversion counts describe the same events? | Gaps over 10% open a tracking row and pause optimization |
| 6 | Lead quality scoring | Score | Score this form fill 0 to 100 against the ICP | Scores are sent back as offline conversion values |
| 7 | Landing page match | Score | Does the page deliver what the ad promised? | Low scores are fixed before spend |
| 8 | Report annotation routing | Noul | Does this weekly change need a client explanation? | Only the yes rows get an LLM-written note |
Every row has a known answer set, which is the test for whether a decision model fits. If the answer has to be written, it belongs to an LLM.
Quick answer: Jev for paid ads means using TypeSafe's decision model for the high-volume judgment steps in paid ads. 1) It answers pick-one, score and yes/no questions for about $0.0004 each. 2) It cannot write, explain or plan. 3) The top fit here is search-term triage (google), then creative tagging (meta, tiktok). 4) Run it behind a confidence threshold, with an approval queue for the middle band. 5) Keep a frontier model for anything a person reads and a human on anything hard to undo. 6) Access is by waitlist as of September 2026.
What Jev is, in one table
Jev is a model from TypeSafe AI, the company founded by Diogo Almeida after his work on ChatGPT at OpenAI. It opened in early access on September 15, 2026. TypeSafe calls it a System One model: you send it a block of text (the state) and a set of typed questions, and it returns typed answers with a probability attached. It does not write sentences.
Question types
- What TypeSafe reports
- Choice (pick one option), Score (a number in a range), Noul (a yes/no probability)
Price
- What TypeSafe reports
- $0.042 per 1M input tokens. Output is free. Roughly $0.0004 per decision
Latency
- What TypeSafe reports
- 70 to 500 ms per call, all answers returned in one parallel pass
Structured-output errors
- What TypeSafe reports
- 0%, because the output is schema-constrained
Accuracy on TypeSafe's 4-workflow benchmark
- What TypeSafe reports
- Jev 67.8%, Claude Opus 5 73.1%, GPT-5.6 Sol 74.1%
Access
- What TypeSafe reports
- Waitlist. Early users on X report approval the same day or the next
What it cannot do
- What TypeSafe reports
- Write text, code or a rationale. It gives a number and no explanation
| Property | What TypeSafe reports |
|---|---|
| Question types | Choice (pick one option), Score (a number in a range), Noul (a yes/no probability) |
| Price | $0.042 per 1M input tokens. Output is free. Roughly $0.0004 per decision |
| Latency | 70 to 500 ms per call, all answers returned in one parallel pass |
| Structured-output errors | 0%, because the output is schema-constrained |
| Accuracy on TypeSafe's 4-workflow benchmark | Jev 67.8%, Claude Opus 5 73.1%, GPT-5.6 Sol 74.1% |
| Access | Waitlist. Early users on X report approval the same day or the next |
| What it cannot do | Write text, code or a rationale. It gives a number and no explanation |
Two caveats belong next to those numbers. The benchmarks are vendor-reported and measure agreement with frontier models, and no large independent reproduction exists yet. And TypeSafe says itself that it cannot prove the launch price is unsubsidized. Treat the price as today's price.
The 8 jobs, one by one
1. Search-term triage (Google)
Question: "Buyer, researcher, job seeker, competitor or junk?" Answer type: Choice. Next: Negatives above the threshold, the rest queued.
The highest-volume job in paid search. Covered in depth in Jev for Google Ads.
2. Creative tagging (Meta, TikTok)
Question: "Hook, format, offer, awareness stage?" Answer type: Choice ×4. Next: Performance grouped by tag.
The job behind the 724-ads demo. See Jev for Meta Ads.
3. Anomaly triage
Question: "Is this CPA jump tracking, auction, creative or budget?" Answer type: Choice. Next: Routed to the right check before anyone is paged.
The numbers come from the platform. The first question in every incident is which kind it is, and that is a four-option Choice.
4. Change safety gate
Question: "Is this proposed change reversible and under the account's limits?" Answer type: Noul. Next: Yes is applied, no waits for approval.
A second opinion on every change an agent proposes. Cheap enough to run on all of them.
5. Tracking sanity
Question: "Do these two conversion counts describe the same events?" Answer type: Noul. Next: Gaps over 10% open a tracking row and pause optimization.
Platform vs analytics vs store orders. The threshold is arithmetic, and the judgment about whether two event names mean the same thing is Jev's.
6. Lead quality scoring
Question: "Score this form fill 0 to 100 against the ICP" Answer type: Score. Next: Scores are sent back as offline conversion values.
Value-based bidding needs a value per lead within hours. A Score per lead at this price makes that routine.
7. Landing page match
Question: "Does the page deliver what the ad promised?" Answer type: Score. Next: Low scores are fixed before spend.
Same question on every channel, so one checker covers Google, Meta and LinkedIn.
8. Report annotation routing
Question: "Does this weekly change need a client explanation?" Answer type: Noul. Next: Only the yes rows get an LLM-written note.
Most rows in a weekly report need no comment. Deciding which ones do is cheap, and writing the comment is the LLM's job.
Free SEO/GEO audit
- Surface technical SEO issues
- Spot quick wins
- Find duplicate and thin pages
Free · no credit card · instant
Ryze AI — Autonomous Marketing
A decision is only useful if something acts on it
- Ryze AI reads the live account and makes the change
- Approval queue for budgets, bids and money pages
- Free to connect, flat fee for autopilot
2,000+
Marketers
$500M+
Ad spend
23
Countries
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.
The decide-then-act loop
- Collect the rows with code or an API: search terms, ads, pages, AI answers. No model needed.
- Shortlist with rules or embeddings, so Jev sees 10 to 15 candidates per item and not the whole account.
- Ask Jev one bounded question per row: a Choice, a Score or a Noul. Every answer comes back with a probability.
- Apply a threshold. High confidence is applied automatically. The middle band goes to an approval queue. Low confidence is dropped.
- Write with an LLM only where text is needed: the new ad, the new title, the one-line reason a client will read.
- Verify the LLM's output with Jev again before it ships.
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.
Search-term triage
- Rows
- 20,000 terms
- Approx. input tokens
- ~1.6M
- Cost at $0.042 per 1M
- ~$0.07
Creative tagging
- Rows
- 5,000 ads
- Approx. input tokens
- ~3M
- Cost at $0.042 per 1M
- ~$0.13
Lead scoring
- Rows
- 3,000 leads
- Approx. input tokens
- ~1.5M
- Cost at $0.042 per 1M
- ~$0.06
| Job | Rows | Approx. input tokens | Cost at $0.042 per 1M |
|---|---|---|---|
| Search-term triage | 20,000 terms | ~1.6M | ~$0.07 |
| Creative tagging | 5,000 ads | ~3M | ~$0.13 |
| Lead scoring | 3,000 leads | ~1.5M | ~$0.06 |
The point of the table is the order of magnitude. Jobs 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, diagnosing a drop once it is routed, and every explanation a client reads. Keep a human on budgets and bids over 20%, new campaigns, and conversion tracking.
Where to go deeper on each job
A decision model only sorts. The guides below cover the work around it: getting the data, writing the rules, and making the changes.
Alex M.
VP Growth
Multi-Brand Ecommerce Group
We had five channels and five people each defending their own budget. Handing allocation to Ryze AI cut blended CAC 27% in a quarter — not because the bidding got smarter, but because money finally moved to whichever channel was winning that week.”
27%
Blended CAC reduction
5
Channels automated
12 weeks
To full autonomy
1,000+ marketers use Ryze
Automating hundreds of agencies
★★★★★How to start before you have access
Jev is in early access. There is no public self-serve signup yet.
1. Join the waitlist
Request access at typesafe.ai. Early users report approval the same day or within a day, but TypeSafe publishes no guarantee.
2. Label a sample first
Before you have a key, hand-label 500 rows of the job you care about. That set is how you will pick thresholds, and it works for any model.
3. Build the loop on a model you can call today
The decide-then-act loop runs on Claude Haiku 4.5 or Gemini Flash-Lite with structured outputs. It costs more per call, and the code does not change when you swap Jev in.
4. Connect the account data
The rows have to come from somewhere. The Ryze AI MCP for Google Ads and Meta Ads gives Claude, ChatGPT, Cursor and Grok live access to the account, so the collect step is one prompt.
The Ryze help center documents AI ad management across Google, Meta, TikTok and LinkedIn and approvals.
