> This page is for version v1.0.1 (default).
> For other versions, use one of these documentation indexes:
> - v1.0.1 (default): https://docs.kinetk.ai/v-1-0-1/llms.txt

> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.kinetk.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.kinetk.ai/_mcp/server.

# Introduction

> Public HTTP API and MCP server for the KINETK Knowledge Graph.

![KINETK turns scattered social posts into a connected map of narratives, creators and patterns](/_fern-img/002ae500d35fe1643d07624e16e59ed1c8ebc219ef96672719bfa50ecaa7e0e1.webp)

The **KINETK Graph Service API** is the multimodal content intelligence platform of the social web. It generates narratives and visual patterns across TikTok, Instagram, Reddit, Pinterest, X, Snapchat and more.

### There are two ways to use it:

* **HTTP API** — REST endpoints behind an API key. Best for backends, dashboards, scheduled jobs and agents you build yourself.
* **MCP server** — a hosted [Model Context Protocol](https://modelcontextprotocol.io) server that exposes the graph as tools in Claude, ChatGPT, Cursor, Codex and other AI assistants.

*Both reach the same backend and bill the same account.*

#### [Get Your API Key](https://platform.kinetk.ai)

Create a key in the KINETK platform to start making requests.

## What you can do

| Capability                                                              | Endpoint                                                                | Speed         | Cost                    |
| ----------------------------------------------------------------------- | ----------------------------------------------------------------------- | ------------- | ----------------------- |
| Ranked content + its narratives and creators, in one call               | `POST /search`                                                          | a few seconds | 0.01 credits per record |
| Trending narratives                                                     | `GET /narratives/trending`                                              | under 2 s     | free                    |
| Search the narrative map                                                | `GET /narratives/search`                                                | under 2 s     | free                    |
| One narrative: posts, creators, platform split                          | `GET /narratives/{id}`                                                  | under 3 s     | free                    |
| A narrative's momentum over time                                        | `GET /narratives/{id}/trajectory`                                       | under 2 s     | free                    |
| Browse the visual pattern map                                           | `GET /intelligence/patterns`                                            | under 2 s     | free                    |
| One pattern, its posts and its lookalikes                               | `GET /intelligence/patterns/{patternKey}` (+ `/content`, `/lookalikes`) | under 2 s     | free                    |
| Deep retrieval and analysis: `records`, `insights`, `patterns`, `stats` | `POST /intelligence/jobs` + `GET /intelligence/jobs/{id}`               | async         | per record or per call  |
| Your credit history                                                     | `GET /usage`                                                            | under 1 s     | free                    |
| Service status and data freshness                                       | `GET /health`                                                           | under 1 s     | free                    |

Three kinds of call sit behind that table:

* **Precomputed reads** of the narrative and pattern maps, rebuilt on a schedule — instant and free.
* **Synchronous search** — one bounded live retrieval, for agents and apps that call many times per task.
* **Async jobs** — deep retrieval and analysis over the live corpus. Submit returns a `jobId` immediately; poll until it succeeds. Identical requests are served from cache.

Not sure which to use? See [Choosing the Right Call](/documentation/guides/choosing-the-right-call).

## Core concepts

* **Content** — a post or video, with normalized relevance, engagement and recency signals and the score it is ranked by. See [The Content Object](/documentation/guides/the-content-object).
* **Narratives** — clusters of content a community is talking about, tracked over time with momentum, lifecycle and sentiment. See [Narrative Intelligence](/documentation/guides/narrative-intelligence).
* **Patterns** — groups of content that *look* alike, discovered from images and video, with the visual elements they share. See [Pattern Intelligence](/documentation/guides/pattern-intelligence).

## Base URL

```
https://api.kinetk.ai/graph
```

Every endpoint requires your secret key in the `x-api-key` header. Export these so the examples in these docs work as written:

```bash
export API_BASE="https://api.kinetk.ai/graph"
export API_KEY="<your key>"
```

## Get started

1. [Quickstart](/documentation/get-started/quickstart) — get an API key, try queries in the Playground, then make your first calls.
2. [Pricing & Credits](/documentation/get-started/pricing-credits) — what each call costs.
3. [Hosted MCP Setup](/documentation/mcp-server/hosted-mcp-setup) — connect your AI assistant.