What teams build on public social data.
Four recurring jobs, and the endpoints each one needs. Build them into a product or pipeline through the REST API, or run them in an AI or automation tool through MCP.
Whichever you start with, it is the same key, the same credits, the same response model, and the same five platforms.
Creator and social analytics products
Creator vetting, dashboards, engagement analysis, social apps.
Report on any account and the content it publishes. Resolve a profile, page its recent posts, and you have follower counts, engagement, and posting cadence in structured responses. Shared concepts map to consistent fields across supported responses, while platform-specific capabilities remain available as documented fields.
- Creator DiscoveryFind the accounts worth vetting, before you have a handle to look up.How it works
- Influencer VettingTurn a handle into a public profile, recent posts, and engagement signals on every platform the creator publishes on.How it works
- Creator BenchmarkingPut a creator’s numbers next to their peers’, on fields that mean the same thing.How it works
- Audience ResearchFind out who shows up around a topic, and what they say when they do.How it works
Social listening and competitive monitoring
Brand mentions, competitor activity, posting patterns, audience response.
Watch accounts and topics you do not control. Platforms are built to be scrolled, not subscribed to — so run the same searches and feeds on a schedule instead, diff the results, and catch a competitor’s move or a mention of your brand the day it happens.
- Competitor MonitoringTrack what competitors post and how fast they grow — without opening their profiles.How it works
- Social ListeningFind what people are posting about your brand or category, then read the discussion underneath.How it works
- Brand MonitoringKeep owned and earned activity in one dataset, on one schedule.How it works
- Reddit Community MonitoringWatch the subreddits your category actually lives in, thread by thread.How it works
AI agents and content intelligence
Current social context, content research, transcripts, comments, trend analysis.
Give a model current social context it cannot reach on its own. Transcripts turn video into text and comment threads turn discussion into text, so an agent can read what people are actually saying instead of counting views — through the REST API inside a product feature, or through MCP when the agent should fetch it itself.
- Get video transcriptTikTok/v1/tiktok/videos/{videoId}/transcripts/{language}
- Get video commentsTikTok/v1/tiktok/videos/{videoId}/comments
- Get post commentsInstagram/v1/instagram/posts/{postId}/comments
- Get post commentsReddit/v1/reddit/posts/{postId}/comments
- Get post video transcriptLinkedIn/v1/linkedin/posts/{postId}/transcript
- Trend DiscoveryFind what is gaining traction in your category, and read what it actually says.How it works
- Content ResearchSee what already works in your category before you commit a production budget.How it works
- Video Content AnalysisHand a model the words in a video, not the view count.How it works
- Comment AnalysisGive a model the discussion under a post, replies and all.How it works
Market intelligence and data pipelines
Company signals, alternative data, enrichment, recurring datasets, quantitative and academic research.
Social data as an input to something that is not a social product: enriching a CRM record from a company URL, adding a public signal to a dataset, or running a recurring extract for research. Batch endpoints take up to 100 ids at a time, so refreshing a whole table is a handful of scheduled calls rather than a queue to manage.
- Lead EnrichmentFill in a CRM record from the URL that is already on it.How it works
- Dataset BuildingCollect once, then refresh the whole table in a handful of calls.How it works
- Talent SourcingPut evidence on a sourcing list, not just a job title.How it works
- Academic ResearchA sampling frame you can define once and re-run for a year.How it works
Two ways to run any of them.
The same product, the same credits, and the same normalized responses. Which one fits depends on whether the job lives inside software you are building or inside a tool your team already uses.
REST API
For a product feature, a scheduled pipeline, or anything that has to run without a person watching. Familiar HTTP, one authentication scheme across every platform, and responses you can parse once and reuse.
MCP
For recurring research inside an AI or automation client you already use. Connect a supported client once and the same endpoints become tools the agent can call — no integration to build, and the same credits.
Start with the one you need.
Get a free API key and make the first call. 100 credits, every platform, no credit card.