TOEFL · 2026 Format

TOEFL 2026 Format: Sections, Task Types, Timing, and Scoring Explained

A complete, section-by-section breakdown of the TOEFL 2026 format — all 12 task types, adaptive structure, timing, and how the new 1-6 scoring scale works across every part of the test.

Reviewed by the LingoLeap Research Team · Updated May 2026 against the April 2026 ETS Test Blueprint

12 task types

Across all 4 sections

~2 hours

Total test time

1–6 scale

Per-section scoring

Quick Answer

What is the TOEFL 2026 format?

TOEFL 2026 has four sections: Reading (~27–30 min, adaptive), Listening (~25–29 min, adaptive), Writing (~23 min, linear), and Speaking (~8 min, linear). The test includes 12 task types in total, several of which are new to this version. Every section is scored on a 1–6 scale. The April 2026 ETS Test Blueprint officially identifies three task types as AI-scored: Write an Email, Write for an Academic Discussion, and Take an Interview. All other task types — including Listen and Repeat and Build a Sentence — are machine-scored against predefined answers.

TOEFL 2026 Format Overview

The table below provides a comprehensive view of all four sections, their task types, approximate timing, testing mode, and scoring.

SectionTimeTask TypesModeScore
Reading~27–30 minComplete the Words, Read in Daily Life, Read an Academic Passage (~50 Qs)Adaptive1–6
Listening~25–29 minChoose a Response, Conversation, Announcement, Academic TalkAdaptive1–6
Writing~23 minBuild a Sentence, Write an Email, Academic DiscussionLinear1–6
Speaking~8 minListen and Repeat (7 items), Take an Interview (4 Qs)Linear1–6

Reading Format

The Reading section runs approximately 27–30 minutes and uses multistage adaptive testing. The April 2026 ETS blueprint specifies 50 total items across three task types (some used to calibrate future test material rather than counting toward your score).

Complete the Words

A vocabulary-in-context task where you select words to complete short passages. Tests recognition of high-frequency academic and everyday vocabulary.

Read in Daily Life

Short, practical texts such as notices, emails, and signs. Tests comprehension of everyday written English in real-world contexts.

Read an Academic Passage

Longer passages from academic subjects. Tests understanding of main ideas, details, inference, and rhetorical structure — similar to the traditional TOEFL reading task.

Reading is delivered as a router module followed by either a lower or upper module — confirmed by the April 2026 ETS blueprint. The router runs about 18–21 minutes; the assigned module adds roughly 9 minutes. Some items are non-scored (used to calibrate future test material), so the exact number of scored items varies. ETS has not publicly disclosed the routing thresholds or the precise psychometric scoring algorithm, so accuracy across the whole section — particularly the router — is the most reliable strategic focus.

Listening Format

The Listening section runs approximately 25–29 minutes (router ~18 min + lower module 7 min or upper module 11 min) and uses the same router/module adaptive structure as Reading. It includes four task types spanning conversational, functional, and academic listening contexts. As with Reading, some items are non-scored and used for calibration.

Choose a Response

A new task type. You hear a short prompt and select the most appropriate spoken response. Tests comprehension of functional language and conversational cues.

Conversation

A dialogue between two speakers in an academic or campus setting. Questions test comprehension of purpose, details, and speaker attitudes.

Announcement

A short monologue delivering practical information such as a class update, event notice, or campus announcement.

Academic Talk

A longer lecture or academic presentation. Tests comprehension of complex ideas, organization, and detail from extended spoken discourse.

Writing Format

The Writing section runs approximately 23 minutes and is linear — the same task sequence for every test-taker. It contains three distinct task types.

Build a Sentence (machine-scored)

10 items. You arrange jumbled words or phrases into grammatically correct sentences. Machine-scored against predefined answers; tests sentence structure, grammar control, and word order.

Write an Email (AI-scored)

1 item. You write a short functional email in response to a scenario. The April 2026 ETS blueprint labels this task as AI-scored, with the model evaluating qualities such as fluency, coherence, grammar, and overall communication effectiveness.

Write for an Academic Discussion (AI-scored)

1 item. You contribute to an online academic discussion. The blueprint labels this task as AI-scored; the model evaluates argument development, vocabulary, grammar, and coherence within an academic register.

Speaking Format

The Speaking section runs approximately 8 minutes and is linear. It has two task types that test very different speaking skills.

Listen and Repeat (7 items, machine-scored)

You listen to a sentence and repeat it as accurately as possible. Per the April 2026 ETS blueprint, this task is machine-scored against predefined answers — testing pronunciation accuracy and listening precision. 7 items total.

Take an Interview (4 items, AI-scored)

You respond to open-ended questions in an interview format. The blueprint identifies this as AI-scored; the model evaluates fluency, coherence, grammar, vocabulary, and communication effectiveness. 4 items total.

Adaptive Testing Explained

Multistage adaptive testing (MST) means a section is delivered in stages, with the difficulty of the later stage selected based on performance in the earlier stage. The April 2026 ETS blueprint confirms this design for Reading and Listening, naming the stages explicitly: a router module followed by either a lower or upper module.

In practical terms:

  • All test-takers begin with the same router module.
  • Stronger router performance appears to route test-takers to the upper module; less consistent performance routes to the lower module. ETS has not publicly disclosed the routing thresholds.
  • The upper module typically presents more challenging items; the lower module presents more accessible ones. ETS has not detailed exactly how router and module performance combine into the final 1–6 band.
  • Some Reading and Listening items are non-scored — they are used to calibrate future test material rather than counting toward your band.

Because the algorithm is undisclosed, the most defensible strategy is consistent accuracy across the section — especially in the router, since it is the only stage every test-taker shares.

Writing and Speaking are not adaptive. The blueprint states explicitly that every test-taker of a specific form receives the same Writing and Speaking tasks.

The 1-6 Scoring Scale

Every section of TOEFL 2026 is scored on a 1-6 scale. This replaced the previous 0-30 per-section system. The scale reflects proficiency levels that align with university admission requirements.

ScoreLevelWhat It Indicates
6AdvancedFull mastery of academic English in this section
5StrongConsistent high performance with minimal gaps
4CompetentHandles most academic tasks effectively
3DevelopingPartial success; noticeable gaps in proficiency
2LimitedSignificant difficulty with academic tasks
1Below MinimumVery limited proficiency demonstrated

Key Format Differences from Previous TOEFL

If you have studied for an older version of the TOEFL, these are the most important format changes to be aware of for TOEFL 2026:

FeatureOld TOEFLTOEFL 2026
Total time~3 hours~2 hours
Scoring0–30 per section1–6 per section
Adaptive testingNot presentReading & Listening
New task typesNone6 new types across all sections
Speaking tasks4 integrated + 2 independentListen & Repeat + Interview

For a full side-by-side comparison of old and new formats, see our TOEFL 2026 vs Old TOEFL guide. TOEFL 2026 vs Old TOEFL guide.

Practice TOEFL 2026 With AI Feedback

Master every task type in the new TOEFL 2026 format with adaptive practice and instant AI scoring from LingoLeap.

Start Free Practice

Frequently Asked Questions

What is the TOEFL 2026 format?

The TOEFL 2026 format consists of four sections: Reading (~27–30 min, adaptive), Listening (~25–29 min, adaptive), Writing (~23 min, linear), and Speaking (~8 min, linear). Each section uses new task types introduced alongside retained formats, and all four are scored on a 1–6 scale.

How many task types are in TOEFL 2026?

TOEFL 2026 has 12 task types across four sections: 3 in Reading (Complete the Words, Read in Daily Life, Read an Academic Passage), 4 in Listening (Choose a Response, Conversation, Announcement, Academic Talk), 3 in Writing (Build a Sentence, Write an Email, Academic Discussion), and 2 in Speaking (Listen and Repeat, Take an Interview).

Which TOEFL 2026 sections are adaptive?

Reading and Listening use multistage adaptive testing — a router module followed by either a lower or upper module — as confirmed in the April 2026 ETS blueprint. Writing and Speaking are linear: every test-taker of a specific form receives the same Writing and Speaking tasks.

How long is TOEFL 2026?

Per the April 2026 ETS Blueprint, working test time runs roughly 1 hr 23 min to 1 hr 29 min: about 27–30 minutes for Reading, 25–29 minutes for Listening, 23 minutes for Writing, and 8 minutes for Speaking. Once instructions and transitions are included, the total appointment is about 2 hours.

How is TOEFL 2026 different from the old TOEFL format?

TOEFL 2026 introduced new task types in every section, replaced the 0-30 per-section scoring with a 1-6 scale, reduced total test time from about 3 hours to 2 hours, and added adaptive testing to Reading and Listening.

Related Guides