Speakonaut uses advanced AI models trained on official exam criteria to evaluate your speaking and writing responses. Our scoring system analyzes multiple aspects of your performance—fluency, vocabulary, grammar, pronunciation, and coherence—just like human examiners do.
Instead of a single number, we show you a score range (e.g., 6.0–7.0) because automated scoring—like human scoring—has inherent variability. The range reflects realistic uncertainty and helps you set appropriate expectations.
Our confidence levels (low, medium, high) indicate how certain the AI is about its assessment. Higher confidence typically comes from clear, well-structured responses. Lower confidence may suggest inconsistent performance or edge cases.
You are graded in real time across four equally weighted criteria. Examiners want natural, clear communication under pressure—not memorised speeches.
Both tasks are graded using four criteria. Task 2 carries double weight, so essay quality strongly influences your final band.
Reading is scored objectively. The challenge is understanding information quickly and accurately across multiple texts and question types.
Listening is also objectively scored. You are tested on accurate understanding while dealing with accents, pace and distractions.
PTE Speaking is scored by AI on ONLY 3 traits: Content, Pronunciation (0-5), and Oral Fluency (0-5). Grammar and Vocabulary are NOT scored separately.
PTE Essay uses 7 traits. Summarize Written Text uses 4 traits. CRITICAL: If Content=0 OR Form=0, the TOTAL SCORE IS 0.
PTE Reading is automatically scored. Some tasks have NEGATIVE MARKING.
PTE Listening includes integrated tasks. Some have NEGATIVE MARKING, others don't.