5 Best Social Media Datasets

updated on Nov 22, 2025

We compared five leading social media data providers, focusing on the types of social data they offer and the platforms they include.

Our evaluation finds vendors fall into two groups: those offering content-level social media data (posts, comments, engagement) and those providing profile- or identity-level data (social handles, professional profiles, company info). See the platform coverage comparison of the best social media dataset services:

Platform	Bright Data	Oxylabs	PDL	Coresignal	Cognism
Instagram	Comments Posts Profiles Reels	❌	Profile links only	Creator metadata only	❌
TikTok	Comments Posts Profiles Shop	❌	❌	❌	❌
YouTube	Comments Profiles Video posts	✅	Profile links	Creator metadata	❌
Facebook	Comments Company Events Posts Profiles	❌	Profile links	❌	❌
Twitter/X	Posts Profiles	❌	Profile links	❌	❌
Reddit	Posts Comments	❌	❌	User profiles	❌
LinkedIn	Posts Profiles Company joblistings	❌	✅	✅	✅
Pinterest	Posts Profiles	❌	❌	❌	❌
Quora	Posts	❌	Profile links	❌	❌
GitHub	Repository	❌	Profile links	Developer profiles	❌

Before you start evaluating individual vendors, it’s helpful to remember that not all social media data providers offer the same types of data. The field is actually split into two clear categories, depending on what the provider delivers.

To simplify the evaluation for readers, we categorize providers into two main groups:

These vendors collect and deliver raw or enriched social media content, including:

Posts (text, media metadata, hashtags, views, likes)
Comments and replies
Engagement metrics (likes, shares, reposts, views)

Providers in this category:

Bright Data
Oxylabs

These providers are suitable for teams involved in AI/ML model training, user sentiment analysis, content analytics, or any application that needs post-level data.

These vendors focus on public profile information, not on posts or comments that may include:

Social media account URLs/handles (LinkedIn, Facebook, Twitter/X, Instagram, GitHub, etc.)
Professional and demographic data
Employment and education history
Company–employee relationship data

Providers in this category:

People Data Labs (PDL)
Coresignal
Cognism

These datasets can be invaluable for a variety of purposes, like enriching your CRM, gaining sales insights, enhancing HR technology, understanding people better through analytics, or connecting profile data with content datasets from other providers.

Bright Data is a leading public web data platform with 31 specialized social media datasets covering major platforms such as Instagram, Facebook, TikTok, LinkedIn, Reddit, Pinterest, Quora, Bluesky, and X (formerly Twitter).

Bright Data’s marketplace indicates three primary data layers. These dataset types appear across platforms such as Instagram, TikTok, LinkedIn, and Reddit.

1. User profiles:

Username/profile name
Bio/description
Followers / following / subscriber counts
Engagement metrics (avg. likes, comments, shares)
Page/business account metadata
Account categories (creator, brand, business, etc.)

2. Posts:

Post text, captions, or titles
Media metadata (image/video content)
Hashtags, mentions, links
View counts, like counts, share counts
Publishing timestamps
Engagement ratios
Topic fields and content categories

Examples from the marketplace include:

Instagram: Posts
X (Twitter): Posts
Facebook: Posts by Profile URL
TikTok: Posts

3. Comments:

Comment text
Commenter profile metadata
Likes/reactions
Thread/reply structure
Comment timestamps
Engagement metrics for discussion activity

Delivery and format

Bulk datasets (CSV, JSON, NDJSON, Parquet)
API endpoints for continuous or real-time pulls
Cloud delivery options for large dataset integrations

Pricing

Dataset-based pricing (one-time or subscription)
API usage-based pricing for ongoing data collection

Oxylabs provides custom datasets for YouTube, including metadata, transcripts, and 720p+ resolution, to support training and fine-tuning AI models. Unlike Bright Data’s marketplace, which offers ready-to-download data, Oxylabs emphasizes on-demand data collection.

1. User profiles

Typically supports the collection of:
Username/display name
Bio/description
Followers, following, subscriber counts
Location fields (when publicly available)
Profile category (creator, business, athlete, entertainer, etc.)
Public URLs, profile links, and external site references

2. Posts and content objects

Typical fields included:

Post text, captions, or titles
Media metadata (image, carousel, thumbnail, video indicators)
View counts, likes counts, and favorites
Hashtags, mentions, tagged profiles
Post URLs and identifiers
Posting timestamps
Engagement rates (calculated or extracted)

3. Comments and discussion data

Using post-level endpoints, Oxylabs retrieves:
Comment text
Comment author name/handle
Reactions, likes, upvotes
Thread/reply depth
Comment timestamps
Comment IDs + parent IDs (thread structure)

Delivery and format

Delivered as CSV, JSON, or Parquet
Stored in client’s S3 / GCS / Azure buckets
Weekly, daily, hourly, or real-time refresh

Pricing

Custom pricing
Often based on platform count, refresh frequency, and dataset size

People Data Labs (PDL) is a provider of social media data, but its focus is limited to profile-level information. Unlike Bright Data or Oxylabs, which supply detailed content data such as posts, comments, engagement, and raw content datasets, PDL does not offer datasets containing posts, comments, videos, photos, threads, likes, or engagement metrics. Instead, PDL specializes in providing social-profile datasets, including:

PDL supports:

LinkedIn
Facebook
Twitter/X
Instagram
GitHub
Quora
Pinterest
YouTube (as a social link on profiles)

Delivery and format

APIs: Person Enrichment API, Person Search API, Bulk Person Enrichment API.
Bulk dataset licenses: Data can be delivered via S3, Snowflake, Azure, GCP, or direct download.
Schema documentation: Available Person Schema, field bundles, and field availability tables.

Pricing

API credit-based pricing.
Bulk dataset licensing: subset datasets (e.g., Email Dataset, Consumer Social Dataset, etc) available under licensed terms.
Free trial: They offer a free tier (e.g., 100 API calls/month) for testing.

Unlike social media data sources that primarily focus on content, Coresignal is dedicated to providing detailed profile-level and organizational data, with limited coverage of platforms like TikTok, Instagram, and Facebook.

Types of data provided

1. User profiles

Coresignal aggregates public user profiles from platforms such as:

Reddit (user profiles, metadata)
GitHub (developer profiles, repository metadata)
StackOverflow (user profiles, activity stats)
Professional networking sites (public employment/education fields)

Typical profile fields include:

Username
Display name
Bio/about section
Profile links
Activity metrics (karma score, commit counts, reputation, etc.)
Location fields (when publicly available)
Skills, technologies, topics of interest

2. Company and organizational data

Coresignal also specializes in:

Company profiles
Employee lists
Funding rounds (when public)
Industry and company categorization
Company–employee graph data

3. Creator and influencer metadata (limited)

Coresignal provides metadata for:

YouTube creators
Instagram creator profiles (public metadata only)

Delivery and format

Coresignal provides data through:

Bulk datasets (JSON, Parquet, CSV)
Continuous data updates (weekly/monthly)
API access (for subsets of data)

Platforms covered

Public social / UGC / tech platforms:

Reddit
GitHub
StackOverflow
Other developer and tech communities

Professional and business websites:

Corporate websites
Company registries
Public business directories

Creator platforms (metadata only):

YouTube
Instagram

No raw content platforms (posts/comments):

TikTok, Facebook, Twitter/X: Not supported for content-level extraction

Pricing model

Dataset licensing (one-time or subscription)
Pricing based on:
- Dataset size
- Fields included
- Update frequency
- Data refresh volume
No usage-based scraping billing (since Coresignal sells data, not scraping requests)

Cognism positions itself as a Software-as-a-Service (SaaS) and data provider, rather than a scraper or a marketplace for datasets. There are no consumer-platform datasets (such as TikTok or Instagram); the focus is solely on professional and work-related identity data.

Types of data provided

1. Professional profiles

While Cognism does not deliver raw social media posts or comments, it does include public social profile URLs, typically most commonly LinkedIn. Cognism keeps an extensive database of business professionals, including:

Full name
Job title & seniority
Employment history
Company affiliation
LinkedIn-style role metadata
Work experience timeline
Skills & industry classification

2. Contact and enrichment data

Cognism’s business model mainly focuses on:

Verified business emails
Business phone numbers (with verification levels)
GDPR-compliant contact data
Territory-based coverage

3. Company data

Cognism provides structured company datasets, such as:

Company size, industry, revenue band
Hiring insights
Technology stack signals
Company growth indicators
Employee counts & org structure

Delivery and format

Unlike Bright Data or Oxylabs, Cognism takes a different approach to data. Instead of selling downloadable datasets of posts or large raw data files, Cognism provides its data through a more tailored, accessible approach that better suits your needs.