> **For coding agents and LLMs:** This is one Social Fetch marketing page (markdown export). For API orientation start with [`/llms.txt`](https://www.socialfetch.dev/llms.txt); for auth and crawl rules use [`/agents.txt`](https://www.socialfetch.dev/agents.txt); for every operation use [`/llms-endpoints.txt`](https://www.socialfetch.dev/llms-endpoints.txt).

## This page

- **On-site (HTML):** <https://www.socialfetch.dev/compare/crowdtangle>
- **Markdown (.mdx) URL:** <https://www.socialfetch.dev/compare/crowdtangle.mdx>

---

# CrowdTangle alternative: Social Fetch vs CrowdTangle

Meta shut CrowdTangle down on August 14, 2024. Its replacement, the Meta Content Library, is free but only open to approved academic and non-profit researchers. Social Fetch is a self-serve API for public Facebook and Instagram data, without CrowdTangle's dashboards, archive or search across all posts.

## At a glance

|  | Social Fetch | CrowdTangle |
| --- | --- | --- |
| Status | Live, self-serve | CrowdTangle shut down August 14, 2024; Meta Content Library replaced it for researchers |
| Who can use it | Anyone who signs up: companies, agencies, newsrooms, developers | Researchers at qualifying academic or not-for-profit research institutions |
| Getting access | Sign up, get 100 free credits (no card), call with one key | Application through Meta's Research Tools Manager, independent review, then Meta processing |
| Price | 1 credit per request: about $0.00165 on the Scale pack | No fees in Meta's Secure Research Environment |
| Search and archive | Lookups by Page, profile, group or post URL; no search across all posts | Searches Meta's public content archive, up to 100,000 results per API query |
| Where the data goes | Your own database, product or notebook | Analysis inside Meta's Secure Research Environment or an approved cleanroom |

## Pricing (last verified 2026-09-24)

**What does public Facebook data cost now?** CrowdTangle can't be bought at any price any more. The Meta Content Library costs nothing if you qualify, but most companies don't. Social Fetch bills 1 credit per completed request for Facebook Pages, posts, comments and groups, and for most Instagram routes, so the question is volume, not eligibility.

|  | Social Fetch | CrowdTangle | Note |
| --- | --- | --- | --- |
| Can you still use CrowdTangle? | Not applicable | CrowdTangle: no, not since August 14, 2024 |  |
| Single Facebook Page lookup | ~$0.00165 | Free, for approved researchers only |  |
| Latest posts from 200 Pages, daily for a month | 6,000 credits: about $10 at the Scale pack rate, or the $29 plan (18,000 credits) | Free, for approved researchers only |  |
| Keyword search across all public Facebook posts | Not available | Yes, inside Meta's environment |  |
| Time to first request | Minutes | About 2 to 3 weeks of review, then 2 to 3 weeks of Meta processing |  |
| Using the data in a product | Yes, subject to your own legal review | Not its purpose: access is for research, and outputs leave under agreed terms |  |

CrowdTangle status from Meta's [CrowdTangle page](https://transparency.meta.com/researchtools/other-data-catalogue/crowdtangle/) (no longer available as of August 14, 2024). Content Library coverage, environments, fees and the 100,000-result query cap from the [Meta Content Library and API docs](https://developers.facebook.com/docs/content-library-and-api); eligibility, application route and review times from [Get access](https://developers.facebook.com/docs/content-library-and-api/get-access). CrowdTangle's former features and audience from [Columbia Journalism Review](https://www.cjr.org/tow_center/meta-is-getting-rid-of-crowdtangle.php) (July 2024); announcement timing from [Nieman Lab](https://www.niemanlab.org/2024/03/a-window-into-facebook-closes-as-meta-sets-a-date-to-shut-down-crowdtangle/) (March 2024). All checked September 24, 2026. Social Fetch figures use the Scale pack rate ($379 for 230,000 credits, about $1.64 per 1,000; September 2026) unless noted; Starter is $14 for 1,000, and plans run $29 for 18,000, $79 for 50,000 and $225 for 150,000 credits a month. Credit costs per route from the [API reference](https://www.socialfetch.dev/docs/api/v1/facebook).

## Feature matrix

| Feature | Social Fetch | CrowdTangle |
| --- | --- | --- |
| Public Facebook Pages, profiles and posts | Yes | Yes |
| Public Facebook groups and group posts | Yes | Yes |
| Comments on public posts | Yes | Yes |
| Instagram accounts, posts and comments | Yes | Yes |
| Facebook events and Marketplace listings | Yes | Yes |
| Keyword search across all public posts | No | Yes |
| Archive of past public content | No | Yes |
| Free for eligible users | No | Yes |
| Open to companies, agencies and newsrooms | Yes | No |
| Self-serve signup, no application | Yes | No |
| Data in your own systems, not a clean room | Yes | Partial |
| Facebook Ad Library search in the same API | Yes | No |
| TikTok, X, YouTube and Reddit on the same key | Yes | No |

## What happened to CrowdTangle

CrowdTangle was Meta's public insights tool for Facebook and Instagram. Newsrooms, researchers, fact-checkers and many others used its dashboards, saved lists, search and alerts to see which public posts were spreading. Meta announced the shutdown in March 2024, and as of August 14, 2024 it is no longer available.

Meta's replacement is the Meta Content Library, a web interface, and the Content Library API. Together they cover public Facebook Pages, posts, profiles, groups, events, comments and Marketplace listings, and public Instagram accounts, posts and comments. They are research tools, not a new CrowdTangle for everyone.

## Who can use the Meta Content Library

Meta's eligibility rules cover two groups: people at academic institutions that are accredited, degree-granting and not-for-profit, and people at not-for-profit institutions whose primary purpose is scientific or public-interest research. You apply through Meta's Research Tools Manager; an independent partner reviews the application (Meta's docs name ICPSR at the University of Michigan and France's CASD), which Meta says takes 2 to 3 weeks, followed by another 2 to 3 weeks of processing.

If you qualify, it's a good deal: there are no fees for the web interface or for the API in Meta's Secure Research Environment, and API queries return up to 100,000 results. If you work at a brand, an agency, a startup or most newsrooms, you don't meet those criteria. Columbia Journalism Review's reporting on the switch made the same point about news organisations.

## The clean-room trade-off

Content Library API analysis runs inside Meta's Secure Research Environment or an approved third-party cleanroom, in Python or R. You can export research outputs under agreed terms, and downloading a public data subset is limited to the web interface for approved applicants. That fits a study. It doesn't fit a product feature, a client dashboard or an alert that has to reach Slack.

Social Fetch returns JSON to your own servers. What you do with it is up to you and your counsel: Meta's terms restrict automated collection without its permission, and Social Fetch can't grant that permission.

## What Social Fetch covers on Facebook and Instagram

On Facebook (21 endpoints): Pages and public profiles by URL (`GET /v1/facebook/profiles`), their posts, reels, photos and events (`/v1/facebook/profiles/posts`), single posts, comments and replies (`/v1/facebook/posts/comments`), video transcripts, public groups and group posts (`/v1/facebook/groups/posts`), events, Marketplace listings and the Facebook Ad Library (`/v1/facebook/ad-library/ads/search`). On Instagram (20 endpoints): profiles, posts, reels, comments, highlights, hashtag search and transcripts. Most routes cost 1 credit.

The same key covers TikTok, X, YouTube, Reddit and the rest of the 25 platforms. Route lists: [Facebook hub](https://www.socialfetch.dev/platforms/facebook), [Instagram hub](https://www.socialfetch.dev/platforms/instagram). For group research specifically, see the [Facebook group data API](https://www.socialfetch.dev/use-cases/facebook-group-data-api) page.

## What Social Fetch can't replace

Be clear about the gaps before you migrate a CrowdTangle workflow. Social Fetch has no keyword search across all public Facebook posts: you look up known Pages, groups and posts by URL (on Instagram, hashtag search covers some discovery). It has no archive: you get what a Page's public feed shows when you call it, not a stored history of past posts and their engagement. There are no dashboards, saved lists, leaderboards or Chrome extension.

For alerts, [Monitors](https://www.socialfetch.dev/monitoring) watch Instagram posts, reels and hashtags and send a signed webhook when something new appears, but Monitors don't cover Facebook yet; for Facebook Pages you schedule the posts call yourself and diff the results. For a wider survey of options, see [best Facebook data APIs in 2026](https://www.socialfetch.dev/blog/best-facebook-data-apis-2026).

## When CrowdTangle is the better choice

- You're a researcher at a qualifying academic or not-for-profit research institution
- You need keyword search across all public Facebook and Instagram content
- You need Meta's archive of past public posts
- Your budget is zero and you're happy to work inside Meta's clean room

## When Social Fetch is the better choice

- You're a brand, agency, startup or newsroom that doesn't qualify for the Content Library
- You need data in your own systems for a product, dashboard or alert
- You track known Pages, groups or accounts and need access today, not in a month
- You also need the Facebook Ad Library, TikTok, X, YouTube or Reddit
- You want webhooks when Instagram accounts post

## Verdict: Qualifying researchers: the Meta Content Library. Everyone else: Social Fetch.

CrowdTangle is gone. If you're an academic or non-profit researcher, apply for the Meta Content Library: it's free, has search and an archive, and is the closest thing to CrowdTangle's data. If you're not eligible, or you need the data in your own product, Social Fetch returns public Facebook Pages, posts, comments and groups, plus Instagram and other platforms, self-serve and per request. It doesn't replace CrowdTangle's search, archive or dashboards.

## FAQ

### Is CrowdTangle still available?

No. Meta says CrowdTangle has been unavailable since August 14, 2024. Existing dashboards and API access ended then.

### What replaced CrowdTangle?

The Meta Content Library and Content Library API. They cover public Facebook Pages, posts, profiles, groups, events, comments and Marketplace listings, and public Instagram accounts, posts and comments.

### Who can access the Meta Content Library?

People at accredited, degree-granting, not-for-profit academic institutions, and at not-for-profit institutions whose primary purpose is scientific or public-interest research. Applications go through Meta's Research Tools Manager and an independent review. Companies and agencies don't meet those criteria.

### Does the Meta Content Library cost anything?

Meta says there are no fees for the web interface or for the API in its Secure Research Environment. The catch is eligibility and the requirement to analyse data inside Meta's environment or an approved cleanroom.

### Can Social Fetch search all public Facebook posts like CrowdTangle?

No. Social Fetch looks up known Pages, profiles, groups and posts by URL, and searches Facebook events, Marketplace and the Ad Library. There is no keyword search across all Facebook posts.

### Can I monitor Facebook Pages with Social Fetch?

Yes, by scheduling `GET /v1/facebook/profiles/posts` for each Page (1 credit per page of results) and comparing results. Monitors with built-in webhooks cover Instagram, TikTok, X and Reddit, but not Facebook yet.

### Is it allowed to collect public Facebook data with a third-party API?

Meta's terms restrict automated collection without its permission, and Social Fetch can't grant that permission. Whether a public-data service fits your use depends on your jurisdiction, the data and how you use it. If you need Meta-sanctioned access and qualify, use the Content Library. This is not legal advice.
