Your SEO data, in your own stack.
A documented REST API over rankings, AI visibility, keywords, audit issues, content drafts, and research sessions. Authenticate with a key, name a site, get JSON.
Included on Pro and Agency plans — not a separate add-on.
The data you already own, over plain HTTPS.
The SearchChamp REST API exposes the same numbers your dashboard shows — as JSON, scoped to the sites in your workspace. It exists for the work that does not happen inside an SEO tool: the internal dashboard, the warehouse sync, the client report, the alert that fires when a page slips. Send a key, name a site, get an answer.
One key, one site, one call.
Create a key under Settings → API keys, send it as a Bearer token, and name the site you are asking about. Every route follows this shape.
curl -sS \ -H "Authorization: Bearer atl_YOUR_KEY" \ "https://api.searchchamp.com/v1/api/rankings?site=example.com"
Parameters, response shapes, and the full endpoint list are in the endpoint reference.
Six surfaces, read and write-back.
Rankings
Rank movers for a site over a window — what gained, what slipped, and the page that ranks for it.
AI visibility
Citation presence across the AI answer engines you track, summarised for a site and a period.
Keywords
Every keyword tracked for a site, with its current position and the metadata you set when you added it.
Audit issues
The latest site-audit summary by severity and category, so a dashboard can trend technical health over time.
Content drafts
The Content Writer feed, including article bodies — and a write-back so a publish is attributed to the right date.
Research sessions
Research Studio sessions and their findings, for pulling a finished brief into a downstream workflow.
Keyed, scoped, and rate limited by plan.
- Authentication is an API key sent as a Bearer token — a web session will not reach this surface. Keys are created and revoked under Settings → API keys.
- Every key is scoped to the organization that issued it, and every request names its site with a ?site= parameter. Asking for a site you do not own returns a 404, never another tenant’s rows.
- Reads are GET and the two write-backs are PATCH. No other verb is accepted — anything else answers 405 rather than being quietly served as a read.
- Read and write are separate permissions: a read-only key cannot reach either write route.
- Calls draw on the same monthly allowance as the rest of your workspace, metered at the same cost as the equivalent MCP tool, so an automation cannot quietly outspend your plan.
Rate limits are counted per minute against your plan: 300 requests a minute on Pro and 600 on Agency. Go over and the API answers 429 and names your limit, so a client can back off from the response instead of guessing.
The reference lives in the Help Center.
Endpoints, parameters, response shapes, error codes, and per-feature coverage are documented in full under API Documentation. It is updated in the same work as any change to the API itself, so it never trails the contract.
Feature coverage
What the API can and cannot reach today, capability by capability.
Read the docs →Authentication
Creating a key, sending it, scoping it, and rotating it when it leaks.
Read the docs →Endpoints
Every route with its parameters, response shape, and a worked example.
Read the docs →Errors and limits
Status codes, error bodies, rate limits, and how a client should back off.
Read the docs →REST API — FAQ
Yes, and deliberately so. The REST routes call the same query cores the dashboard and the MCP tools call, so there is exactly one implementation per capability and no chance of the API and the UI disagreeing about your numbers.
Use the REST API when a program consumes the data: a dashboard, a sync job, an internal tool. Use the MCP connector when a person is asking questions inside an AI app and wants the answer in the conversation. They read the same data; they are different front doors. See /resources/mcp for the connector.
In the Help Center, under API Documentation. It carries the endpoint list, parameters, response shapes, error codes, and per-feature coverage — and it is updated in the same work as any change to the API itself, so it never trails the contract.
Mostly read. Two write-backs let you report a publish — one for a Content Writer article, one for a Research Studio session — and both need a key with write access. Everything else is read-only.
Neither today. It is plain HTTPS with a Bearer token and JSON responses, so any HTTP client works, and polling on a schedule is the current pattern for keeping an external system in sync. Nothing you build now would need rebuilding if a client library ships later.
These routes return scoped results for a site and a window, not a full dump of your account. For a one-off migration or a complete historical export, contact us at hello@searchchamp.com rather than paginating the API for a week.
Pro and Agency. A key issued on a lower plan authenticates, but the API routes answer 403 until the workspace is on Pro or above — the API is included in those plans rather than sold as a separate line item.
Requests are counted per minute against your plan: 300 a minute on Pro and 600 on Agency. Exceeding it returns a 429 that names your limit, so a client can back off from the response rather than guess.
Yes. A call is metered at the same cost as the equivalent MCP tool, so neither surface is a cheaper way to pull the same data. It draws on the workspace allowance you already have rather than being billed separately.
Issue a key and make your first call.
API keys live under Settings → API keys in your SearchChamp workspace. The full reference is in the Help Center.