21 full-length mocks through Santa AI by Socra AI (formerly Riiid, South Korea), an official ETS partner. Speaking and Writing are scored by ETS's own SpeechRater® and e-rater® engines — the same technology that marks the real exam.
Most TOEFL simulators write their own questions and their own scoring. That is fine for drilling, but the band it hands you is an approximation of an approximation — their guess at how ETS would mark an item ETS never wrote.
These mocks are the other way round: 21full-length tests on ETS-licensed content, with Speaking and Writing sent to ETS's own engines. The score you practise against is produced the way the real one will be.
That matters more in 2026 than it did in 2025, because the scale changed. TOEFL now reports 1–6 bands in half steps rather than 0–120, and a simulator still calibrated to the old scale is telling you something the exam no longer says.
All of it for ₹7,500.
EEC students get 21 full-length TOEFL mock tests on ETS-licensed content through Santa AI, with unlimited retakes for six months. ETS also publishes free sample questions on its own site — useful for a first look, but not full scored simulations.
ETS publishes free practice sets officially, and EEC's free demo class includes a scored sample section so you can see the 2026 format and the AI scoring before enrolling. The full-length scored mocks come with the course.
Santa AI TOEFL is the practice platform built by Socra AI (formerly Riiid, South Korea) — an official ETS partner — ETS-licensed content, official ETS scoring (SpeechRater® and e-rater®). Speaking and Writing are scored by ETS's own SpeechRater® and e-rater® engines, the same technology behind the real test.
Yes. TOEFL iBT changed on 21 January 2026 — shorter, adaptive Reading and Listening, and a 1–6 band scale in 0.5 steps instead of 0–120. Every mock here runs the new format, and score reports show the legacy 0–120 equivalent as ETS does through 2028.
Six to ten full-length mocks is typical before test day, spaced far enough apart to fix what each one exposes. Retakes are unlimited, so the number is set by when your score stops moving, not by how many the package allows.
More reliably than most, because these are ETS-licensed questions marked by ETS's own scoring engines rather than a third party's approximation. Treat a single mock as a diagnosis and the trend across three as the prediction.