TOEFL AI Tutor by TOEFLGRADER
How we score TOEFL Writing
We teach what raises the official task scores. We do not invent a /30 scale or borrow descriptors from other English tests.
Machine items
Build a Sentence is exact chunk order.
Constructed responses
Dual-model 0–5 scoring with Score-0 gates.
Indicative section map
1.0 + 5 × mean(BAS/10, Email/5, AD/5), rounded to 0.5. This is orientation only.
How it works
1
Read a sample report
See /sample-report
Frequently asked questions
Practice scores follow public descriptors. They cannot guarantee a test-day result.
Try TOEFL AI Tutor free
Get criterion feedback, track progress, and follow a personalized plan toward your target band.
See a sample report