Pricing
Pay for the runs, not for the seat
Every plan does the same four-stage analysis with the same fabrication boundary. What changes is how much of it you can do — metered in tokens, measured from what the models actually consumed, at about 20,000 for a full run.
Enough to tailor a résumé against a posting properly, and to see whether the gap list tells you anything you didn't know.
- Token allowance
- 60Kper month
- JD analyses
- 3per month
- Résumés rendered
- 2per analysis
- Full answer drafts
- 3per month
- Course matches
- 2per gap
- Roadmaps built
- 1per month
- Job searches
- Offper month
For an actual search — several postings a week, all eleven templates, a worked answer for every question, a roadmap on every analysis, and job search.
- Token allowance
- 800Kper month
- JD analyses
- 40per month
- Résumés rendered
- 11per analysis
- Full answer drafts
- 40per month
- Course matches
- 4per gap
- Roadmaps built
- 40per month
- Job searches
- 60per month
For coaches and career centres running many searches at once, where the meter is the thing that actually binds.
- Token allowance
- 3Mper month
- JD analyses
- 150per month
- Résumés rendered
- 11per analysis
- Full answer drafts
- 200per month
- Course matches
- 6per gap
- Roadmaps built
- 150per month
- Job searches
- 200per month
What a token is
The unit the pipeline actually consumes, measured from every model call rather than estimated after the fact. We show it rather than hiding it behind a credit.
- Reading the posting3,600
- Matching against your profile4,800
- Rewriting your résumé8,400
- Preparing questions and courses3,200
- A full analysis20,000
Coverage, gaps and question frameworks cost nothing. A run that cannot be afforded is refused before its first stage, never half-run.
If you run out mid-cycle
One-off packs, on any paid plan. They never renew and they never expire — a pack bought in the last week of a cycle is not a partial purchase, because anything unspent carries into the next one. The plan allowance itself does not carry; only these do.
Side by side
Every row is a number enforced at its own seam. Nothing here is a feature list — if it is on this table, there is a constraint behind it.
What happens when you reach one
A cap refuses the next new thing and never takes away something you already have. Each of these is enforced at its own seam, which is why running out of one does not read as running out of everything.
Measured from what the models actually consumed — a full run is about 20,000. When it empties we stop before the first stage rather than half-run an analysis and bill you for it.
Analyses you have already run stay readable at any cap.
Below eleven is fewer drafts, never worse ones — highest-ATS first.
Every question keeps its framework and its source bullets, at every cap including zero.
The gap is always shown, with or without a course beside it.
A roadmap you have built stays readable and tickable at any cap.
Saved jobs and their statuses stay at any cap, including Off.