Instagram, transcripts

Instagram reels to text, in bulk, billed by the minute

Turn on includeTranscript and every video in the run comes back with a transcript and its detected language, in the same row as the caption, play count, likes, comments and the creator's follower count. A 30-second reel costs about $0.0059 all-in: $0.0019 for the row plus $0.004 for one started minute of audio.

• $0.004 per started minute
• $1.90 per 1,000 rows
• Any language, up to 15 minutes
• No login

Cost calculator

Enter how many reels you want and their typical length. The table uses each tool's Free-plan price as listed on Apify on 2026-10-15.

We are not the lowest price in every case. apple_yang/instagram-transcripts-scraper is a fraction of a cent lower for reels under a minute on the Free plan, and we come out slightly lower at three minutes. What you get here is the transcript plus the complete post record, Instagram's own AI summary and the latest comments, from any source the Actor reads, in one run. The transcript minute is $0.004 on every Apify plan, and the row itself drops from $0.0019 to $0.0015 on Scale and $0.0011 on Business.

What's in each row

• transcript and transcriptLanguage: the spoken words and the detected language
• caption, hashtags, mentions: what the creator wrote
• likeCount, commentCount, video.playCount, video.duration: how it performed and how long it is
• owner: username, full name, verified flag and follower count, so you can normalise performance by audience size
• engagementRate: likes plus comments divided by followers, as a percentage
• aiTitle and aiSummary with includeAiSummary: Instagram's own AI description of the post
• latestComments with commentsPerPost: up to 200 of the newest comments
• audio, createdAt, url, isPaidPartnership and the rest of the standard post fields

Where the reels come from

How transcription works, and where it falls short

• Speech-to-text on the reel's audio, in any language, for videos up to 15 minutes long.
• Billed per started minute at $0.004, only when transcripts are turned on.
• No speech, usually no transcript. Reels with no detectable speech, such as reels set only to a licensed song, are usually skipped for transcription. Now and then a reel still comes back with an empty transcript and its minute billed, so check the transcript column on large runs.
• Speech only. Text shown on screen or burned into the video isn't read.
• Plain text. The transcript is a single text field with no timestamps or SRT file.
• Machine-made. Names, slang and loud background music cause mistakes. Check a transcript before you quote it.

What people do with reel transcripts

• Hook research. Pull the last 100 reels from the top creators in a niche and compare their first sentences with their play counts.
• Repurposing. Turn your own reels into blog posts, newsletters or captions for other platforms.
• Search by what was said. Captions are often a line and a pile of hashtags. Transcripts let you find reels about a product or claim even when the caption doesn't mention it.
• Creator briefs. Summarise how a creator talks about a category before you pitch them, together with the public engagement signals.
• LLM pipelines. Each row is already structured text plus numbers, ready for summarisation, tagging or embeddings.

Use it as an API

There is no separate transcript endpoint to learn: you call the Actor through Apify's API with includeTranscript: true and read the rows. Small jobs can use the one-call endpoint below, which runs the Actor and returns the rows within 5 minutes. For bigger batches, start a run, then read its dataset when it finishes, asking only for the columns you need:

curl -X POST "https://api.apify.com/v2/acts/yugenox~instagram-scraper/runs?maxTotalChargeUsd=2" \
  -H "Authorization: Bearer $APIFY_TOKEN" -H "Content-Type: application/json" \
  -d '{"startUrls":["https://www.instagram.com/nasa/reels/"],"maxItems":200,"includeTranscript":true}'

curl "https://api.apify.com/v2/datasets/<defaultDatasetId>/items?fields=url,transcriptLanguage,transcript,video" \
  -H "Authorization: Bearer $APIFY_TOKEN"

maxTotalChargeUsd caps what the run may cost, here $2 for at most 200 reels. The run object returned by the first call has the defaultDatasetId, and a webhook can tell your server when the run is done. The Python and JavaScript clients below do the waiting for you. In an AI assistant, the same call is one sentence: see the MCP setup.

Run it from code

The last 30 reels from a profile's reels tab, with transcripts and AI summaries. At 30 seconds each that is about $0.18.

// Save as run.mjs, then: npm i apify-client && APIFY_TOKEN=... node run.mjs
import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });

const run = await client.actor('yugenox/instagram-scraper').call({
  startUrls: [
    "https://www.instagram.com/nasa/reels/"
  ],
  maxItems: 30,
  includeTranscript: true,
  includeAiSummary: true
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) console.log(item.url, item.transcriptLanguage, item.transcript);

import os
from apify_client import ApifyClient  # pip install apify-client

client = ApifyClient(os.environ["APIFY_TOKEN"])

run = client.actor("yugenox/instagram-scraper").call(run_input={
    "startUrls": ["https://www.instagram.com/nasa/reels/"],
    "maxItems": 30,
    "includeTranscript": True,
    "includeAiSummary": True,
})
if run is None:
    raise RuntimeError("Run did not finish")

for item in client.dataset(run.default_dataset_id).iterate_items():
    print(item.get("url"), item.get("transcriptLanguage"), item.get("transcript"))

# Small jobs: runs the Actor and returns the rows in one call (up to 5 minutes).
curl -X POST "https://api.apify.com/v2/acts/yugenox~instagram-scraper/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"startUrls":["https://www.instagram.com/nasa/reels/"],"maxItems":30,"includeTranscript":true,"includeAiSummary":true}'

Last updated 2026-10-06.