Developer docs

The AIMentionTracker API, CLI and MCP server

Read your brand's AI visibility from code: how often each assistant names you, the answers behind every number, and which sources those engines read first. Six read-only endpoints, one header, no SDK required.

Start 7-day free trialStart with the quickstart

Card required, nothing charged for 7 days

What you get from the AI visibility API

One API. Three ways to reach it.

Everything runs on a read-only REST API at https://app.aimentiontracker.ai/api/v1, authenticated with an organisation key you mint at API keys in the app.

The CLI and the MCP server are conveniences over that same API. Neither can do anything the API cannot, and all three obey the same rules below.

Your first request
curl -H "X-API-Key: $AMT_API_KEY" \
  https://app.aimentiontracker.ai/api/v1/me
The API keys screen in the AIMentionTracker app, listing existing organisation keys with their scopes. No key value is shown; keys are displayed once at creation and stored hashed.
Mint a key yourself, in seconds. The value is shown once and stored as a hash — which is why there is no key visible here.

Read these before you render a number

We sell measurement honesty, so the API is opinionated about how its output may be presented. A client that ignores the three rules below will produce a report that is confidently wrong, in your customer's name, and it will look fine until somebody checks it.

Three rules you must not break

These are not style preferences. Breaking them produces a confidently wrong report in your customer's name.

  1. 1

    A null rate means "not enough answers yet". It is NOT zero.

    Rates come back as {present, responses, rate, min_sample}. When rate is null, the sample is below the floor. Say “3 of 4 answers, too few to quote a percentage”. Never render it as 0%, and never let it into an average. Reporting “0% visibility” for a brand that was named in 3 of 4 answers is the single most damaging thing you can do with this data.

  2. 2

    Never average the engines.

    There is no single visibility score and you must not compute one. ChatGPT, Perplexity and Google AI Overviews are different surfaces that disagree, so a mean describes nothing real. Report per engine. If the user insists on one number, give them the pooled counts (present and responses summed) and say plainly what you did.

  3. 3

    An empty alerts list does NOT mean nothing changed.

    A finding is raised only when both comparison windows carry enough answers AND the move is larger than its own margin of error. On a small plan that excludes most real movement. Say “no change large enough to distinguish from noise at this sample size”. GET /me shows the thresholds in force.

The long version, including why a failed collection on our side never counts as your brand being absent, is on how to read our numbers.

The six endpoints

All GET. All read-only. The full reference has parameters and returned fields.

Every page in these docs

Common questions about the API

Answered from the implementation.

Does AIMentionTracker have an API?
Yes. A read-only REST API at app.aimentiontracker.ai/api/v1, authenticated with an organisation key you mint in the app. There are six GET endpoints covering your brands, per-engine visibility, the stored answers, the sources engines cite and your alerts. A CLI and an MCP server sit on top of the same API.
Is the AIMentionTracker API read-only?
Yes, deliberately, in v1. There are no write endpoints because the two obvious ones both spend money without a human present: adding a prompt expands a metered bill, and triggering a run spends vendor credit. Those stay in the app where a person confirms them.
How do I get an API key?
Mint one in the app under API keys. Owner and Admin roles only. The key is shown once, stored as a SHA-256 hash, and cannot be recovered afterwards. Keys belong to the organisation rather than to a person, so they keep working when someone leaves.
What is the rate limit on the AIMentionTracker API?
300 requests a minute, the same on every plan including the cheapest. Every response carries X-RateLimit-Limit, X-RateLimit-Remaining and X-RateLimit-Reset, and a 429 adds Retry-After so a client knows exactly when to try again.
Does AIMentionTracker have an MCP server?
Yes. @aimentiontracker/mcp exposes six read-only tools to any MCP client, over a hosted HTTP transport or a local stdio process. The hosted container holds no secrets of its own: your key arrives in the header and is passed through to the same v1 API.
Which language should I use?
Any. It is plain HTTP with a header, so curl, fetch, requests or an HTTP client in whatever you already use will do. The CLI exists for terminals and CI, and the MCP server exists for AI assistants. None of them can do anything the API cannot.

7-day trial

Measure it first, automate it second

The API reads what we have already collected. Add your brand and the prompts your buyers type, and the first answers are readable within minutes.

Start 7-day free trialHow we measure

Card required, nothing charged for 7 days.

On every plan

  • Every answer stored in full, with its citations.
  • Every rate shown with the number of checks behind it.
  • Cited-but-not-named reported as its own state.