> **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/use-cases/social-data-api-for-developers>
- **Markdown (.mdx) URL:** <https://www.socialfetch.dev/use-cases/social-data-api-for-developers.mdx>

---

# Social data API for developers

**Is there one API for all the major social media platforms?** Social Fetch is one. A single API key covers TikTok, Instagram, YouTube, X, LinkedIn, Reddit and more, and every response has the same envelope: `data`, plus `meta` with the request ID and `creditsCharged`. There is a typed TypeScript SDK, an OpenAPI spec and an MCP server, and you pay per call.

One API key and one response shape for TikTok, Instagram, YouTube, X, LinkedIn, Reddit and more. A typed SDK, an OpenAPI spec, an MCP server, and the credits each call used in every response.

## A real run, measured and priced

One account, three networks, one response shape, run on 21 Nov 2026 against live public data. Three calls with one key, and every response carried the same envelope: the data, plus a meta block with the request ID and credits charged.

| Request | Result | Credits | Time |
| --- | --- | --- | --- |
| `GET /v1/tiktok/profiles/nasa` | Found: 2,211,981 followers and 70 videos. | 1 | 1.8 s |
| `GET /v2/instagram/profiles/nasa` | Found: 104,277,920 followers and 5,018 posts. | 1 | 1.6 s |
| `GET /v1/youtube/channel?handle=NASA` | Found: 15,100,000 subscribers and 6,205 videos. | 1 | 2.2 s |

Total: 3 credits, 5.6 s of server time. Starter pack ($14 per 1,000 credits): $0.042. Scale pack ($1.64 per 1,000 credits): $0.0049.

Every platform's own API is different: its own auth, its own app review, its own rate limits and its own field names. Some have no public API at all, and home-made scrapers break when the page changes.

Five APIs, five of everything. Gluing five of them together means five clients, five error formats and five bills.

You need one client: the same auth header, the same envelope, the same error codes and one balance, whichever network the data comes from.

## Endpoints

| Method | Path | What it does |
| --- | --- | --- |
| GET | `/v1/tiktok/profiles/{handle}` | A TikTok profile: followers, likes, video count and bio. |
| GET | `/v2/instagram/profiles/{profile}` | An Instagram profile by handle, URL or id: followers, post count and bio links. |
| GET | `/v1/youtube/channel` | A YouTube channel by handle, id or URL: subscribers, videos and views. |
| GET | `/v1/twitter/profiles/{handle}` | An X profile with follower and post counts. |
| GET | `/v2/linkedin/profiles` | A LinkedIn person record: roles, schools and followers. |
| POST | `/v1/ask` | Describe a lookup in plain English and it runs the matching route, for prototyping. |

## Workflow

1. **Get a key.** Sign up, copy your API key and send it in the `x-api-key` header on every request.
2. **Install the SDK or use the spec.** The [TypeScript SDK](https://www.socialfetch.dev/docs/sdk) types every route. In other languages, generate a client from the [OpenAPI spec](https://www.socialfetch.dev/openapi.json) or call the REST routes directly.
3. **Branch on `lookupStatus`.** A lookup that finds nothing answers `not_found` rather than failing, so a missing account doesn't look like an outage.
4. **Handle errors once.** Errors share one shape and one set of codes on every route. [Errors](https://www.socialfetch.dev/docs/errors) lists them, with which ones to retry.
5. **Watch your spend.** Each response's `meta.creditsCharged` says what the call cost. Log it with your own job id.

## Example: one profile lookup

```bash
curl -sS \
  -H "x-api-key: YOUR_API_KEY" \
  "https://api.socialfetch.dev/v1/tiktok/profiles/nasa"
```

## Why developers pick Social Fetch

- 87 platforms on one key. No app review per platform and no OAuth dance.
- One envelope. `data` and `meta` on every route, and `lookupStatus` on lookups, so one parser covers them all.
- No requests-per-second cap on paid calls. Your balance is the limit, with a cap on requests in flight; [rate limits](https://www.socialfetch.dev/docs/rate-limits) has the details.
- Pay per call. Looking up one account on TikTok, Instagram and YouTube is 3 credits.
- You only pay for completed lookups. Requests that fail on our side cost nothing.

## FAQ

### Is there one API for all the major social media platforms?

Social Fetch covers TikTok, Instagram, YouTube, X, LinkedIn, Reddit and more with one API key and one JSON envelope, and comes with a TypeScript SDK, an OpenAPI spec and an MCP server.

### Do I need each platform's developer approval?

No. Social Fetch returns public data with your Social Fetch key, so there is no app review on each platform.

### What happens when an account doesn't exist?

The call returns a normal response with `lookupStatus: not_found`, so your code can branch on it instead of catching an error.

### Is there an SDK?

Yes: the [TypeScript SDK](https://www.socialfetch.dev/docs/sdk), on npm as `@socialfetch/sdk`. Calls cost the same credits as the raw API.
