IELTS, TOEFL, ACTFL, and CEFR Speaking Frameworks Compared

Executive finding

CEFR and ACTFL are broad proficiency frameworks; IELTS Speaking and TOEFL iBT Speaking are test-specific score systems. They overlap in the ability to describe language performance, but their levels and scores are not interchangeable labels. Use the official framework tied to the learner's goal, and use only the test provider's researched alignment when comparing systems.

As of 21 January 2026, TOEFL iBT reports an updated 1–6 scale in half-band increments, with comparable 0–120 overall scores during a two-year transition. Any page still presenting only the former 0–30 speaking scale as the current system needs an update.

Comparison at a glance

System	What it is	Main output	Best used for
CEFR	Council of Europe reference framework	A1, A2, B1, B2, C1, C2 plus activity/quality descriptors	Curriculum, can-do goals, assessment interpretation, plurilingual education
ACTFL	Functional proficiency guidelines	Novice through Distinguished, with sublevels in major bands	Describing unrehearsed functional ability across contexts
IELTS Speaking	Live test with an examiner	Band 0–9, including half bands in reporting	IELTS admission, migration, or professional requirements
TOEFL iBT Speaking	Internet-delivered test section	1–6 from Jan 2026; transition comparison to prior scale	TOEFL admission and institutional requirements

This table does not create a conversion. It distinguishes purposes.

CEFR speaking

The Council of Europe describes six proficiency levels and provides can-do descriptors across a large range of language activities. Speaking is not one single score: the framework distinguishes spoken production and spoken interaction and includes qualitative dimensions. The 2020 Companion Volume also expands areas such as mediation and online interaction.

CEFR is a framework of reference, not one centrally administered examination. Different exams and institutions need documented procedures to relate their scores to it. A course calling a unit “B2” does not by itself prove that every learner or task has been externally validated at B2.

Official source and Companion Volume: https://www.coe.int/en/web/common-european-framework-reference-languages/documents

ACTFL Speaking Guidelines 2024

ACTFL's revised 2024 guidelines describe what a person can do with language in unrehearsed, spontaneous contexts. The speaking framework defines proficiency through the FACT criteria:

• Functions and tasks the speaker can accomplish;
• Accuracy and acceptability of the message;
• Context and content in which performance can be sustained;
• Text type produced, from words and phrases through connected discourse.

Major levels include Novice, Intermediate, Advanced, Superior, and Distinguished; the first three are commonly divided into Low, Mid, and High. ACTFL levels should not be converted to CEFR by visual similarity or a homemade table.

Official overview and guidelines: https://www.actfl.org/proficiency-guidelines-overview https://www.actfl.org/proficiency-guidelines-overview

IELTS Speaking

IELTS Speaking is a face-to-face speaking test with three parts. The public band descriptors score four equally important criteria:

• fluency and coherence;
• lexical resource;
• grammatical range and accuracy;
• pronunciation.

The public descriptors define features across bands, but a self-rating from the PDF is not an official test result. The test samples several task types under standardised conditions, and trained rating procedures matter.

Official public descriptors: https://ielts.org/cdn/ielts-guides/ielts-speaking-band-descriptors.pdf

TOEFL iBT Speaking after the 2026 scale update

ETS introduced a 1–6 TOEFL iBT score scale in half-band increments in January 2026. The total is the rounded average of the four section scores rather than a sum. ETS's institutional guide maps the new scale to CEFR and provides comparison ranges to former 0–30 section and 0–120 total scores. It also says reports will show CEFR level, the new overall and section scores, and a comparable old-scale overall score during the two-year transition.

Do not collapse the provider's ranges into a false exact conversion. For example, a new-scale point may correspond to a range of prior speaking scores, and an old total can be produced by different section combinations.

Official scale guide: https://www.ets.org/toefl/institutions/ibt/score-scale-update.html

Why conversion tables disagree

Frameworks differ in:

• purpose and target population;
• task and response format;
• monologue versus interaction;
• construct coverage;
• level boundaries and reporting precision;
• scoring model and rater process;
• evidence used to establish alignment;
• consequences and score-use policy.

An alignment study estimates relationships between systems; it does not make the underlying constructs identical. Institutions should set requirements based on the language demands of their programme and the provider's current score-use research, not copy a third-party SEO table.

Which system should a learner use?

Taking IELTS? Use the current IELTS Speaking format, official public band descriptors, timed practice, and qualified feedback.

Taking TOEFL? Use the current 1–6 scale, current task information, ETS rubrics, and the institution's updated requirement. Do not prepare from pages that assume the old score report is the only one.

Planning a European curriculum or broad can-do pathway? Use CEFR activity descriptors and adapt them to learner context and language.

Describing functional spontaneous ability in an ACTFL context? Use the 2024 ACTFL guidelines and appropriate assessment procedures.

Tracking personal weekly practice? Use task-specific evidence—completed functions, intelligibility, pauses, and repairs. Do not award yourself an official band from an app session.

Rules for product and content teams

1. Name the framework version and access date.
2. Link to the official descriptor or provider page.
3. Separate an estimated practice profile from an official test score.
4. Do not promise an IELTS, TOEFL, CEFR, or ACTFL level from an unvalidated automated score.
5. Preserve ranges and uncertainty in cross-framework alignment.
6. Update old TOEFL scale explanations and redirect duplicate score pages.
7. Have trained assessment specialists review preparation and comparison pages.
8. Never imply endorsement by a test or standards organisation without written authorisation.

Editorial disclosure

This comparison summarises official public documents and does not issue scores, equivalences, or test-preparation guarantees. An accredited language-assessment specialist should verify it before publication, especially after any provider format or score-policy change. All trademarks belong to their respective owners.