Which AI assistants work with the MCP server?

Any client that supports remote MCP servers over HTTP, including Claude Code, Claude Desktop, Cursor and VS Code. Setup takes one command or a few lines of config.

Can I change data through the API?

No. The API and MCP server are read-only. You manage brands, prompts and competitors in the Gensiv app.

Does the API include the full text of AI answers?

No. Each response includes whether your brand was mentioned, its position and sentiment, the competitors named and the sources cited. You can read full answers in the Gensiv dashboard.

What data can I read?

Brands, competitors, prompts with their topics and tags, visibility metrics in total, per day, per prompt and per group, cited domains and URLs, individual AI responses, alerts, Reddit opportunities with subreddit rules, action plans, articles, AI referral traffic and your credit balance.

Is there a Looker Studio connector?

It is in progress. Until it is live, the daily metrics endpoint returns one flat row per day, engine and company, which loads straight into a spreadsheet, data warehouse or BI tool, and every table exports as CSV.

Your AI visibility data, wherever you work

Read the same visibility, share of voice, source and answer data you see in Gensiv from your own code, or ask an AI assistant about it through the MCP server. One API key, every brand in your workspace.

Get your API key

CCadence/MCP serverLast 90 daysDemo data

Which prompts are we losing on ChatGPT, and who wins them?

Used gensiv · get_prompt_metrics

Two prompts need attention over the last 30 days on ChatGPT:

"Project management tool with built-in time tracking": Cadence appears in 38% of answers, down 4 points. Tasklane and Planwise are named most.
"Best free project management app for remote teams": 21% visibility, with Tasklane recommended first in most answers.

Want me to pull the answers where Tasklane was recommended and Cadence was not?

Answers captured from 6 AI engines, in every market you sell in.

ChatGPTLive
PerplexityLive
Google AI OverviewsLive
Google AI ModeLive
GeminiLive
ClaudeLive

API & MCP Server

What can you do with the Gensiv API and MCP server?

The Gensiv API is a read-only REST API for the AI visibility data Gensiv collects: metrics per brand, competitor, day, engine, prompt and topic, the sources AI engines cite, every individual answer, alerts, action plans, articles, Reddit opportunities and AI referral traffic. The MCP server exposes the same data to AI assistants such as Claude and Cursor, so you can ask questions instead of writing queries.

Read-only REST endpoints: 20
MCP tools for AI assistants: 19
Requests per API key: 120/min
Every Pro tier and Enterprise: Every plan

Features

Built for dashboards, agents and reports

The numbers from your Gensiv dashboard, in the tools your team already uses.

Which prompts are we losing on ChatGPT, and who wins them?

Used gensiv · get_prompt_metrics

Two prompts need attention over the last 30 days on ChatGPT:

"Project management tool with built-in time tracking": Cadence appears in 38% of answers, down 4 points. Tasklane and Planwise are named most.
"Best free project management app for remote teams": 21% visibility, with Tasklane recommended first in most answers.

Want me to pull the answers where Tasklane was recommended and Cadence was not?

Ask your AI assistant

Connect Claude, Cursor or VS Code and ask which prompts you are losing, who wins them and which sources they cite.

API keysCreate key

Name	Created	Last used
Looker Studio reportgsk_4f9aQ2…	Sep 2	Today
Claude MCPgsk_b21cT8…	Sep 14	Today
Warehouse syncgsk_7d0eL5…	Aug 21	Yesterday

Keys are shown once, stored hashed, and can be deleted at any time.

Keys you control

Workspace keys, shown once, stored hashed, deleted in one click.

Request

curl "https://api.gensiv.com/v1/brands/$BRAND_ID/metrics?engine=chatgpt" \
  -H "Authorization: Bearer $GENSIV_API_KEY"

200 OK · application/json

{
  "from": "2026-08-29",
  "to": "2026-09-27",
  "responses": 1512,
  "data": [
    {
      "entityType": "brand",
      "name": "Cadence",
      "visibility": 41,
      "shareOfVoice": 25,
      "avgPosition": 2.4,
      "sentiment": 78
    },
    {
      "entityType": "competitor",
      "name": "Tasklane",
      "visibility": 47,
      "shareOfVoice": 29,
      "avgPosition": 1.8,
      "sentiment": 71
    }
  ]
}

A clean REST API

JSON responses, cursor pagination, clear error codes and a full OpenAPI spec.

No code? Export instead

Any table as CSV and any dashboard as a PDF, with your date range and filters applied.

Product tour

From API key to answers

Demo data for Cadence, a fictional project management brand.

CCadence/SettingsLast 90 daysDemo data

API keysCreate key

Name	Created	Last used
Looker Studio reportgsk_4f9aQ2…	Sep 2	Today
Claude MCPgsk_b21cT8…	Sep 14	Today
Warehouse syncgsk_7d0eL5…	Aug 21	Yesterday

Keys are shown once, stored hashed, and can be deleted at any time.

How we measure

Designed to be safe to connect

Your data leaves Gensiv only through keys you create, and only as much as you ask for.

01
Read-only
The API and MCP server can read your data but never change brands, prompts or settings.
02
Scoped to one workspace
A key reads every brand in its workspace and nothing else.
03
Rate limited
Each key can make 120 requests a minute, with clear headers and retry times.
04
Documented
Every endpoint, field and metric formula is in the developer docs.

Related features

AI Visibility AnalyticsHow often AI names you, your rank and sentiment Citation TrackingEvery page AI cites, and whether any are yours Prompt TrackingThe buyer questions you track, and every answer

Frequently asked questions

Last updated October 2, 2026

Every Pro tier and Enterprise. Workspace owners and admins create keys in Settings → API keys.