F
Frame
Reading the question for what it actually asks.
Digital SAT
Knowing what you got wrong is the easy part.
Forge watches how you work the test (where you rush, misread, or second-guess) and turns it into what to change on the next question. The diagnosis and the fix, not just another score.
Built on the method of an instructor whose students break 700 on Reading & Writing four times out of five, most clearing 1500+ overall.
Your reasoning profilesample
What the map measures
Five dimensions, one read.
Every question gets read across five moves of thinking. Together they spell the name, and they pinpoint where your points actually leak.
O
Observe
Catching the details that matter in a passage or problem.
R
Reason
Working through an inference one honest step at a time.
G
Gather
Pulling information together across two places or sources.
E
Execute
Carrying out your approach cleanly, without slips.
What you actually get
It doesn't just score you. It tells you how you test.
After the diagnostic, Forge hands back observations like these, specific to how you (not the average student) move through the test.
Your Forge readdiagnostic · sample
FRAME
78
OBSERVE
66
REASON
71
GATHER
52
EXECUTE
74
What Forge noticedYou hold details well, but lose them when the answer needs two parts of the passage combined. Gather is where your points are leaking, and where your practice starts.Details fine. Combining two parts of passage: points leak. Gather weakest. Practice starts there.
You read accurately, but lose the forest for a tree. On "main purpose" questions you lock onto a vivid detail instead of the author's overall move. Your misses cluster where the right answer is the least flashy option.Read fine. Miss forest for tree. Flashy answer on main-purpose questions: trap. Boring answer usually right.
You answer the question you expected, not the one on screen. Add a constraint like EXCEPT or primarily and your accuracy roughly halves. The content isn't harder; the framing changed.Answer question you expected, not question asked. EXCEPT and primarily cut accuracy in half. Reread the ask.
You're sharpest in the first eight questions, then you speed up and slip. But your second-look changes are right four times out of five, so the instinct to slow down is correct.Start sharp. Speed up. Slip. Second look right four of five times. Slow down.
Illustrative of the read. Your real observations come from your own diagnostic. Caveman mode is on the real report too, one tap.Sample read. Real one comes from your test. Caveman mode on real report too.
How this is different
Chatbots explain. Forge observes.
A chatbot tutor
Explains the question to you.
Helps with one problem at a time.
Starts fresh every conversation.
Forge
Observes how you approached it.
Learns your patterns across sessions.
Remembers, and compounds.
And next to the usual options
| Forge | Khan Academy | AI chatbot | Private tutor | |
|---|---|---|---|---|
| Watches how you work, not just what you got wrong | Yes | No | No | Sometimes |
| Remembers you and compounds across sessions | Yes | Tracks progress | Forgets each chat | Sometimes |
| Steers practice to your specific weak spots | Yes | Somewhat | No | By hand |
| Tells you when it isn't sure yet | Yes | — | No (confidently wrong) | Yes |
| Cost | Free in early access | Free | $0–20/mo | $50–200/hr |
Features and pricing as of 2026; alternatives vary.
Practice that compounds
The longer you practice, the better it knows you.
The profile isn't fixed. You practice; Forge refines it. Within a few weeks it steers practice toward the dimensions you actually need, not the generic weak-spot list everyone gets. That longitudinal record is the moat: it can't be cloned, because it's you. Time to results varies.
Day 1
First read
Week 2
Patterns emerge
1 month
Targeted practice
Know where you stand
See your score. Then watch it climb.
Every student wants the number. Forge gives you one on day one from your diagnostic, then keeps it honest. Log every real SAT, upload your Bluebook and paper practice tests, and your whole trajectory lines up against your goal.
Your score trajectorysample
▲ +180 since your diagnostic · goal 1400
+ Log a score↑ Upload a practice test
Log every score
Real SATs, school PSATs, Bluebook attempts: one timeline, measured against the target you set.Upload a practice test
Snap a photo or PDF of any provider's score report. Forge reads it, sorts each skill into your five dimensions, and folds it into your read. It keeps the signal, never the questions.Scored the way the exam is
Your diagnostic uses Rasch / IRT scaling, the psychometric family behind standardized tests, so it reflects your ability, not which form you drew. A full-length, adaptive practice test on this scale is built in, for when you’ve done the work to make the number mean something.
About the scores. Forge scores are statistical estimates from your responses, to help you track progress. They are not official scores and do not predict or guarantee any result on test day. When you upload a practice test, Forge extracts only your per-skill performance and maps it to your dimensions; the file is processed and not stored, and the questions and passages are never retained. SAT® is a trademark of the College Board and Bluebook™ is its product; the College Board is not affiliated with, and does not endorse, Forge or Vindica Inc.
Start the diagnostic. See your initial profile.
You take the test, you get your full report. No card, no paywall — the results are the whole point.
Start your diagnosticPricing
Free while Forge is in early access.
The early-access deal
Free
No card, no trial clock, no locked features. The price is one finished diagnostic — you give Forge an honest read on how you think, and everything on the right is yours. That includes a learning profile that follows you into whatever Vindica builds next, so you never start a new test from zero. — Kevin
What’s included
- The full, test-condition Digital SAT diagnostic — Rasch-scored, with a real score projection
- Your cognitive map across all five dimensions: the diagnosis, not just a number
- Targeted practice from 5,000+ original & licensed questions, steered to your weak spots
- Your score trajectory, tracked and projected after every session
- The free-forever tools: score-report reader, study-plan generator, and the blog
- A portable learning profile you carry to future Vindica products — AP, TOEFL, and beyond
- Your data, yours to delete in one click, anytime
Who builds this
4 in 5 of founder Kevin Choi’s students break 700 on Reading & Writing.
Most clear 1500+ overall. Kevin is an SAT instructor and the author of the Irreverent Guide to the Digital SAT; Forge is his read on how a student works, made always-on, over a bank of 5,000+ original and licensed questions. That track record is his teaching, not a promise about your score, but it’s the method Forge is built on.
Original & licensed
Rasch / IRT scored
Your data, deletable anytime
Questions
Asked and answered.
Most "AI for the SAT" is a chatbot with a system prompt: it explains the question in front of you and forgets you the second you close the tab. Forge doesn't explain the test. It watches how you work while you take it, and turns that into a profile of how you actually think. Anyone can clone a wrapper in a weekend; nobody can clone a year of your own habits. And it won't tell you you're doing great when you're not.
A quick Google or email sign-in, yes, so your profile saves and builds from the first question. Everything is free while Forge is in early access — no card, nothing to unlock. Want a look before you sign in? The score-report reader needs no account.
It's yours. Deletable and user-controlled: one click, no support ticket. Privacy is built into how the product works.
Your profile updates after every session. Score gains vary by student, and we don't promise numbers we can't keep.