forge.
BETA · JUN 2026
Digital SAT · Cognitive profiling

The first cognitive profiling tool for the SAT.

Forge observes how you think. Answer questions, and a behavioral model of your mind takes shape — sharper than any tutor, because it has observed you, not talked to you.

Full mock test · free during beta
What the map measures

Five dimensions. Locked.

Every question is read across five dimensions of thinking. Together, they spell the name.

F

Frame

Reading the question for what it actually asks.

O

Observe

Catching the details that matter in a passage or problem.

R

Reason

Working through inference one step at a time.

G

Gather

Pulling information together across sources.

E

Execute

Carrying out your approach without slips.

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 thousands.
Remembers, and compounds.
Practice that compounds

The longer you practice, the better it knows you.

The profile isn't fixed. You practice; Seneca refines it. By month two it's targeting the dimensions you actually need, not the same generic weak-spot list everyone gets. The longitudinal record is the moat. It can't be cloned, because it's you. Time to results will vary.

Day 1
First read
Week 2
Patterns emerge
1 month
Targeted practice

Sit a full mock test. See your initial profile.

Full mock test · free during beta
Pricing

Free during the beta.

Beta
$0

Free during beta. No credit card to begin. Pricing after beta, to be announced.

  • The full diagnostic (R&W; Math coming soon)
  • Daily practice sets
  • Your cognitive profile, live
  • Export your data anytime
Who builds this

Built by instructors, not LLMs.

Research-grade behavioral modeling, developed with classroom teachers who have watched students think for years. Not a wrapper. Not a chatbot.

Questions

Asked and answered.

Most “AI for the SAT” is a chatbot with a system prompt. It explains the question in front of you, agrees with whatever you’re leaning toward, and forgets you the second you close the tab. Fine for getting unstuck on one problem. Useless for seeing the pattern across a thousand. Forge doesn’t explain the test to you — it watches how you work while you take it. Where you stall, when you change a right answer to a wrong one, how your reading falls apart in the last five questions. That becomes a profile of how you actually think, and it gets sharper every session. Anyone can clone a wrapper in a weekend; it’s just someone else’s model wearing a prompt. Nobody can clone a year of your own habits. And it won’t tell you you’re doing great when you’re not. When Forge isn’t sure about something yet, it says so.

No. Start with the diagnostic; the account forms around it once you've seen your profile.

It's yours. Exportable, deletable, user-controlled. Privacy is a feature here, built into how the product works.

Your profile updates after every session. Score gains vary by student — we don't promise numbers we can't keep.