Trust through transparency
How assessment works at IELTS Studio
Every number the platform shows you is either a raw score, a published conversion, or a clearly labelled estimate. This page explains each one.
Listening and Reading: raw scores and published anchors
Each correct answer is worth one mark. Scores out of 40 are converted to the 9-band scale using the average anchors published by IELTS: for Listening and Academic Reading, Band 5 is around 16 and 15 marks respectively, Band 6 around 23, Band 7 around 30 and Band 8 around 35. General Training Reading requires more correct answers for the same band (Band 4 is around 15, Band 5 around 23, Band 6 around 30 and Band 7 around 35).
The official guidance notes that the precise mark needed varies slightly from test version to test version. For that reason, partial-module practice reports an indicative band scaled to 40 and says so explicitly, while a completed 40-question module reports an estimated band from the full conversion.
Writing: the four published criteria
Writing responses are assessed against Task Achievement (Task 1) or Task Response (Task 2), Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy. Task 2 carries twice the weight of Task 1 in the real test; in IELTS Studio each task is assessed and reported individually, with the same criteria.
- • Each criterion receives a whole or half band, with comments grounded in the published descriptors.
- • The overall task band is the average of the four criteria, rounded using the official rules.
- • Every strength and improvement quotes short excerpts from your own response — never invented evidence.
- • Under-length responses cap Task Achievement/Response, and this is stated in the feedback.
Speaking: criteria and pronunciation
Speaking is assessed on Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation, each weighted equally. The audio is analysed directly for intelligibility, stress, rhythm and connected speech. Hesitation, repetition and self-correction are assessed by their effect on communication — their presence alone is never penalised.
Overall band: the official rounding rules
The overall band is the average of the four module bands, rounded to the nearest half band. Averages ending in .25 round up to the next half band; averages ending in .75 round up to the next whole band. Mock results apply exactly this arithmetic to the module estimates available.
AI estimates, clearly labelled
Writing and Speaking bands are produced by AI models calibrated to the published criteria. They are practice estimates — never official results — and are labelled as such wherever they appear. No AI system can replace a certified IELTS examiner.
For reliability, writing assessments are routed across two providers. The routing rule is deterministic and server-side: the first two assessed responses each day use OpenAI, and the third and later responses use DeepSeek. Speaking assessment uses Google Gemini's audio understanding. Your responses are validated against a strict schema; malformed output is repaired or retried, and failures are surfaced honestly rather than guessed.
Privacy of your work
Only the content required for assessment is sent to AI providers: your essay text and the task prompt for Writing; your recording and the questions asked for Speaking. Names, email addresses, payment details and unrelated profile information are never sent. Providers process the content to return an assessment and are not permitted to use it for advertising.
What we do not claim
- • We do not claim affiliation with, endorsement by, or authorisation from the IELTS partners.
- • We do not reproduce copyrighted test material. All content is original.
- • We do not present practice estimates as official IELTS scores.
For official format, scoring and registration details, always refer to ielts.org.
Questions about our methodology? Read the privacy policy or contact support from the footer.