instagram-scraper

Fetch public Instagram profiles, posts and media, no login required. TypeScript, zero dependencies (native fetch), built-in retries and rate limiting so you don't get your IP blocked in the first minute.

Requires Node 18 or newer.

Install

npm install @aduptive/instagram-scraper

Usage

Recent posts of a profile

import { InstagramScraper } from '@aduptive/instagram-scraper';

const scraper = new InstagramScraper();
const results = await scraper.getPosts('instagram', 12);

if (results.success && results.posts) {
  console.log(`got ${results.posts.length} posts`);
  await scraper.saveToJson(results, 'posts.json');
} else {
  console.error(results.error);
}

Each post comes with id, shortcode, caption, like/comment counts, timestamp, post URL and the media items (images, videos and carousel children, with dimensions).

Profile info

const result = await scraper.getProfile('instagram');

if (result.success && result.profile) {
  const p = result.profile;
  console.log(p.full_name, p.followers, p.following, p.posts_count);
  console.log(p.biography, p.is_verified, p.profile_pic_url);
}

A single post or reel by URL

const result = await scraper.getPost('https://www.instagram.com/reel/CxSEjxfyJtN');
// also accepts /p/ and /tv/ URLs, or a bare shortcode

if (result.success && result.post) {
  console.log(result.post.media_type, result.post.media_items);
}

Progress and cancellation

getPosts waits 1-2s between posts, so longer runs take a while. You can watch progress and cancel mid-run:

const controller = new AbortController();

const results = await scraper.getPosts('instagram', 12, {
  signal: controller.signal,
  onProgress: ({ fetched, total, currentPost }) => {
    console.log(`${fetched}/${total} ${currentPost.shortcode}`);
  },
});

On failure or cancellation, results.posts still contains whatever was collected before it stopped (with success: false and code: 'ABORTED' when cancelled).

Config

All optional:

const scraper = new InstagramScraper({
  maxRetries: 3,           // retries for network/timeout/5xx errors
  minDelay: 1000,          // random pause between requests, in ms
  maxDelay: 3000,
  timeout: 10000,          // per-request timeout in ms
  rateLimitPerMinute: 30,  // hard cap on requests per minute
});

Notes on how these behave:

• maxRetries is the number of total attempts, and only transient errors (network, timeout, Instagram 5xx) are retried. A 429, 404 or 403 fails immediately, since retrying those just digs the hole deeper. A 200 with a non-JSON body (login wall, HTML error page) fails immediately too, as PARSE_ERROR.
• Getting rate limited (429) or blocked (403) mid-collection stops the run: getPosts returns success: false with the reason and whatever posts were collected so far. It never keeps hammering a blocked IP.
• The rate limit is a sliding one-minute window over every request the scraper makes (profile + media), on top of the random min/max delay.

Error handling

Methods don't throw for scraping failures. Check success:

const results = await scraper.getPosts('someuser');
if (!results.success) {
  // results.error is a message, results.code is one of:
  // RATE_LIMITED, PROFILE_NOT_FOUND, POST_NOT_FOUND, ACCESS_DENIED,
  // TIMEOUT, ABORTED, NETWORK_ERROR, SERVER_ERROR
  console.error(results.code, results.error);
}

Limitations

• Public profiles only.
• Without login Instagram only serves the ~12 most recent posts of a profile, so a higher limit won't get you more than that.
• This relies on Instagram's web API, which they can change whenever they feel like it. If something breaks, open an issue.

Scraping multiple profiles

Do it sequentially, and give it some breathing room between accounts. A shared scraper instance keeps the rate-limit window across profiles:

const scraper = new InstagramScraper();

for (const username of ['nasa', 'natgeo']) {
  const results = await scraper.getPosts(username, 12);
  console.log(username, results.success ? results.posts?.length : results.error);
  await new Promise((r) => setTimeout(r, 60_000));
}

Running scrapes in parallel against Instagram is the fastest way to get rate limited. Don't.

Changed in 2.1

• Media details are now fetched with the post's numeric media id (the correct identifier for that endpoint), with a one-time shortcode fallback.
• A 429/403 during media enrichment now stops the run and reports it, instead of silently returning media-less posts as success.
• Carousels are classified from the post's structural fields, and when the media-info call fails the media already present in the profile response is used as a fallback.
• An already-aborted AbortSignal is respected before any request is made.
• Invalid JSON responses report PARSE_ERROR (no retry) instead of NETWORK_ERROR with retries.
• The constructor validates its config (INVALID_CONFIG error on nonsense like rateLimitPerMinute: 0), and getPosts rejects a non-positive limit.

Migrating from 1.x

• Node 18+ is now required (axios was replaced by native fetch).
• getPosts on failure now also returns the partially collected posts (before, posts was absent on failure).
• Everything else is backwards compatible; getProfile, getPost, signal and onProgress are new.

License

MIT. Use it responsibly and check Instagram's terms of service. This is for collecting public data, not for abuse.