TOEFL Speaking: Take an Interview — points, diagnostics, then the 1–6 band

On the 2026 TOEFL iBT, Speaking is not a vague “band guess” after a long monologue.

Take an Interview is four items, five raw points each. You can see points. You can see diagnostics. The 1–6 Speaking section score still comes from a full sitting — not from one friendly afternoon with the mic unmuted.

Acedify is built for that split: score the tasks you practised, show delivery and language diagnostics you can act on, and keep the section band honest as a range across sittings.

Same philosophy as Know your score before test day. Format depth lives in the TOEFL iBT 2026 complete guide.

Two task types — different failure modes#

If your only Speaking practice is reading model answers aloud, you are not sitting Take an Interview.

---

Points vs diagnostics vs section band#

Keep the three layers separate:

1. Raw points on a task — what happened in that attempt (e.g. interview items out of five each).
2. Diagnostics — fluency, pauses, pronunciation, prosody, language-use notes, words flagged in the transcript.
3. 1–6 section score / overall — average of four sections on the current scale; evidence accumulates across sittings as a range.

Anyone collapsing those three into one theatrical “you’re a 4.5” after a single untimed clip is guessing and hiding it.

---

What good Speaking feedback lists#

After a paid Speaking attempt you should be able to answer:

• Did I lose points on elaboration or on delivery?
• Which words were flagged — and did I say them again wrong on the next item?
• Were my pauses silent gaps or filled hesitation?
• Is my estimate narrowing, or was today one lucky interview?

That is the same fluent / fragile / guessed / gap idea used on timed Reading and Listening — applied to speech.

---

How this sits next to adaptive R/L#

Reading and Listening on the 2026 iBT route adaptively. Speaking does not get “easier modules” because you smiled at the mic — it gets scored against the task.

So the honest prep split is:

• R/L: adaptive timed sittings + explanations on the deciding line
• Speaking/Writing: task rubric + diagnostics
• Overall: range across full evidence — see the guide for 1–6 averaging and the transition 0–120 reporting window

IELTS candidates get the same premium Speaking idea in a different exam language — IELTS Speaking feedback you can act on.

---

Free vs Pro#

Free: untimed practice across sections with explanations — learn task shapes.

Pro: timed sittings, scored Speaking and Writing, adaptive R/L routing, ability estimate as a range, delivery diagnostics and transcript flags.

Microphone required for Speaking. Laptop or phone mic is enough. Nothing uploads until you finish a response; you can delete a recording.

---

A Speaking-focused week (no fake 1–6 lifts)#

1. Untimed Listen and Repeat — study mismatches in the transcript.
2. Timed Take an Interview set — four items, full discipline.
3. Read points + diagnostics; pick one delivery gap and one language gap.
4. Drill; sit interview again.
5. When R/L are also in play, add one adaptive timed module so the overall range uses all four skills.

We will not invent “+0.5 in seven days.” We will insist that unscored interview practice is not a 1–6 prediction.

---

Start with the interview you will actually sit#

Open TOEFL on Acedify. Practise Listen and Repeat and Take an Interview. On Pro, read points and diagnostics — then drill what the transcript flags.

Browse the TOEFL hub →

Next step: Start practicing free on the TOEFL hub. Add scored Speaking when you need feedback you can act on.