How We Measure
Every Number on This Site, Sourced
AccelaStudy® AI publishes specific counts — questions, latency, courses — because a vague claim is easy to make and hard to trust. This page says where each number comes from, how recently it was measured, and what we deliberately do not claim yet.
Question Inventory
1,557,154
The total count of synthesized practice questions across every complete live package on the network, measured by an internal fleet content audit dated July 27, 2026. Every question in that count was adversarially validated and entailment-gated before it shipped — this is a count of finished, quality-passed items, not a target or a draft total.
Adaptive-Update Latency
under 2 milliseconds
How long the AVIAN engine takes to fold a single answer into a learner’s proficiency model and select what to show next. This figure comes from internal engine benchmarks run against the production inference path, not a marketing estimate — it is the same update loop every learner runs on every question.
Course Coverage, by Product
Since the 2026-07-29 catalog split, this site holds no course catalog of its own — every course lives on exactly one product subdomain. The table below is that partition: how many complete, live courses each product currently has, plus the family total.
| Product | Live courses |
|---|---|
| Certs | 161 |
| Free | 150 |
| AP | 27 |
| IB Diploma Programme | 55 |
| Test Prep (SAT, PSAT, ACT, GRE, GMAT) | 11 |
| MCAT — launching October 2026 | 0 |
| LSAT — launching August 2026 | 0 |
| English — launching August 2026 | 0 |
| Languages — launching August 2026 | 0 |
| Family total | 400+ |
GED and USA Citizenship are single-domain free products and are counted in the family total but not broken out as separate rows above. Products with zero live courses are still in content synthesis — the count is genuinely zero, not rounded down.
Study-Time Savings
We do not publish a single percentage for how much study time the adaptive engine saves. It varies by learner — how much relevant prior knowledge you bring to a course, how the course is structured, and how consistently you study — and collapsing that into one headline number would overstate the case for some learners and understate it for others.
What the engine does concretely: it stops presenting concepts you have already demonstrated mastery of, so no learner repeats material they do not need. Rigorous, cohort-based efficacy reporting — tracking real learner outcomes against a baseline, not a marketing estimate — is planned. This section will be updated with methodology and results once that data exists.
Last updated: August 4, 2026. Questions about how a number here was measured? Email press@accelastudy.ai.