Audiogram API
Sign in Get started
← All recipes Recipe 03 · Language teaching

Build an authentic listening lesson from one podcast.

Give learners a real voice, a focused listening task, and transcript evidence, while keeping level choice, explanation, and final teaching decisions with the instructor.

View the prompt ↓ See the example
For
Language tutors, schools and course creators
You provide
A target language, topic and teacher-selected level
Audiogram finds
A published episode and its available transcript
You receive
A teacher-review listening lesson with source evidence
How it works

Turn authentic speech into a structured lesson.

Audiogram supplies the published source and transcript. The instructor chooses the learning goal and checks every activity before class.

Input
[language][topic][teacher-selected level]
  1. 01

    Choose a suitable episode

    Start with one recent, relevant, published source.

    Search
  2. 02

    Retrieve and segment

    Use the available transcript and verify the exact listening passage.

    Transcript
  3. 03

    Build the lesson

    Move from prediction and gist to detail, language in context, and discussion.

    Teacher review
Output

Authentic listening lesson segment · questions · language · answer key

Suggested prompt

Use the transcript as evidence, not as an automatic curriculum.

The instructor chooses the level and suitability. The prompt makes uncertainty and teacher review part of the workflow.

Copy into your agent 
Create a teacher-reviewed listening lesson from one available podcast transcript for [target language] learners at approximately [teacher-selected level].

Use Audiogram to:
1. Find one recent, suitable Apple Podcasts episode about [topic].
2. Retrieve the available transcript.
3. Select one short segment with exact transcript start and end quotes.
4. Return:
   - one pre-listening prediction task;
   - one gist question;
   - four detail questions with transcript evidence;
   - up to five useful expressions in their original context;
   - an answer key and one post-listening discussion or mediation task.

State the episode and transcript availability. Flag unclear speaker labels, missing context, or uncertain wording. Do not automatically assign a CEFR or ACTFL level, invent a transcript, score pronunciation, or call the lesson suitable without teacher review.
Verified example

Short-news listening lab from NPR News Now.

Audiogram returned the real 1,522-character episode transcript. Four exact listening segments were placed with perfect boundary matches.

Four source segments

01 · 15.8 seconds

Negotiations and uncertainty

Gist, attribution and language for uncertain developments.

02 · 14.9 seconds

Cause and consequence in public health

Listening for a chain of reasons and results.

03 · 52.5 seconds

A detainee returns to Washington

Who, what, where, why and reported speech.

04 · 22.9 seconds

How the potato became an insult

Contrast, prediction and cultural vocabulary.

All four boundaries matched exactly. Transcript accuracy, topic sensitivity, language level, and classroom suitability remain teacher decisions.

Lesson snapshot · Segment 01
Before listening
From the headline, predict whether negotiations are moving forward or becoming less certain.
Gist
What is the main change in the possibility of talks?
Listen for detail
Which phrase describes the US effort, and when does Iran say talks could happen?
Language in context
on-again-off-again, come to a deal, and the contrast marker but.
Three follow-up tasks
  1. Retell the update in two sentences without copying the original wording.
  2. Compare how the report attributes the two sides' positions.
  3. Write a neutral headline that preserves the uncertainty in the segment.
Before you teach it

The instructor remains responsible for the lesson.

Does Audiogram assign the learner's level?

No. The instructor chooses the approximate level and adapts the tasks. This recipe does not verify automatic CEFR or ACTFL grading.

Does every podcast have a usable transcript?

No. Episode discovery, transcript availability, accuracy, speaker attribution, and audio quality vary.

Does this score pronunciation or learner answers?

No. Pronunciation scoring, recordings, marking, learner accounts, and progress tracking are outside this verified workflow.

What should be checked before class?

Review the audio and transcript, content sensitivity, language difficulty, cultural context, questions, and answer key.

Use this recipe

Start with one short episode and one learning goal.

Connect Audiogram to a supported agent, paste the prompt, and review the complete lesson before sharing it with learners.

Connect Audiogram → More recipes