Lontiq API

Introduction

Lontiq transforms your company data into a structured, traversable knowledge graph. Documents, spreadsheets, and operational data are automatically organized into entities and relationships that your AI systems can query directly.

The Lontiq API gives you programmatic access to your knowledge graph. You can run Cypher queries, perform semantic search over your entities, retrieve the full graph schema, and chat with your data through an agentic AI loop.

Quick Start

Get up and running in three steps: create an API key, make your first request, and explore your knowledge graph.

1. Get your API key

API keys are created from the Settings page in the Lontiq dashboard by an admin user. When a key is created, the full key is shown once — copy it and store it securely.

Keys follow the format lntq_<40 characters> and can be scoped to specific spaces and permissions.

2. Make your first request

Retrieve your graph schema to see what data is available:

curl -X GET https://api.lontiq.com/api/v1/graph/schema \
  -H "X-API-Key: lntq_your_api_key_here"

import requests

response = requests.get(
    "https://api.lontiq.com/api/v1/graph/schema",
    headers={"X-API-Key": "lntq_your_api_key_here"}
)

schema = response.json()
print(schema["data"])

3. Explore your data

Once you have the schema, you can run Cypher queries to traverse relationships, use semantic search to find entities by meaning, or chat with your graph through the AI-powered chat endpoint.

Authentication

All API requests require authentication via an API key. Pass the key in the X-API-Key header or as a Bearer token in the Authorization header.

# Option 1: X-API-Key header (recommended)
X-API-Key: lntq_your_api_key_here

# Option 2: Bearer token
Authorization: Bearer lntq_your_api_key_here

API key properties

Base URL

All API v1 endpoints are served under:

Append the endpoint path to this base URL. For example, the schema endpoint is at /api/v1/graph/schema.

Get Graph Schema

Returns the full schema of your knowledge graph, including all node labels, relationship types, and their properties. Use this to understand your data structure before running queries.

Query parameters

Required permission

Example request

curl -X GET https://api.lontiq.com/api/v1/graph/schema?format=full \
  -H "X-API-Key: lntq_your_api_key_here"

{
  "status": "Success",
  "data": {
    "node_labels": [
      {
        "label": "Employee",
        "count": 142,
        "properties": [
          { "name": "name", "type": "string" },
          { "name": "department", "type": "string" }
        ]
      }
    ],
    "relationship_types": [
      {
        "type": "REPORTS_TO",
        "count": 138,
        "source": "Employee",
        "target": "Employee"
      }
    ]
  },
  "message": null
}

Query Graph

Execute a read-only Cypher query against your knowledge graph. Write operations (CREATE, MERGE, DELETE, SET, REMOVE) are blocked.

Request body

Required permission

Example request

curl -X POST https://api.lontiq.com/api/v1/graph/query \
  -H "X-API-Key: lntq_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "cypher": "MATCH (e:Employee)-[:REPORTS_TO]->(m:Employee) RETURN e.name, m.name LIMIT 10",
    "params": {}
  }'

import requests

response = requests.post(
    "https://api.lontiq.com/api/v1/graph/query",
    headers={
        "X-API-Key": "lntq_your_api_key_here",
        "Content-Type": "application/json"
    },
    json={
        "cypher": "MATCH (e:Employee) RETURN e.name LIMIT 5",
        "params": {}
    }
)

data = response.json()
for row in data["data"]["results"]:
    print(row)

{
  "status": "Success",
  "data": {
    "results": [
      { "e.name": "Alice Chen", "m.name": "Bob Smith" },
      { "e.name": "Carol Davis", "m.name": "Bob Smith" }
    ],
    "row_count": 2,
    "truncated": false
  },
  "message": null
}

Semantic Search

Search your knowledge graph using natural language. Lontiq uses vector embeddings to find the most relevant entities and their connections, returning nodes, edges, and source references.

Request body

Required permission

Example request

curl -X POST https://api.lontiq.com/api/v1/graph/search \
  -H "X-API-Key: lntq_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{"query": "leadership effectiveness", "limit": 5}'

import requests

response = requests.post(
    "https://api.lontiq.com/api/v1/graph/search",
    headers={
        "X-API-Key": "lntq_your_api_key_here",
        "Content-Type": "application/json"
    },
    json={
        "query": "leadership effectiveness",
        "limit": 5
    }
)

results = response.json()
for node in results["data"]["nodes"]:
    print(node)

{
  "status": "Success",
  "data": {
    "nodes": [
      {
        "id": "4521",
        "labels": ["Competency"],
        "properties": {
          "name": "Leadership Effectiveness",
          "description": "Ability to guide and influence teams..."
        }
      }
    ],
    "edges": [
      {
        "source": "4521",
        "target": "4530",
        "type": "MEASURED_BY"
      }
    ],
    "entry_nodes": ["4521"],
    "sources": ["survey_2024.xlsx"]
  },
  "message": null
}

Chat

Ask questions about your data in natural language. Lontiq uses an agentic AI loop that automatically queries the graph, searches for context, and synthesizes an answer with full source attribution and an audit trail of every tool call.

Request body

Required permission

Example request

curl -X POST https://api.lontiq.com/api/v1/chat \
  -H "X-API-Key: lntq_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{"question": "Who are the top performers in the engineering team?"}'

import requests

# Start a new conversation
response = requests.post(
    "https://api.lontiq.com/api/v1/chat",
    headers={
        "X-API-Key": "lntq_your_api_key_here",
        "Content-Type": "application/json"
    },
    json={"question": "Who are the top performers in engineering?"}
)

result = response.json()
print(result["data"]["answer"])

# Continue the conversation
session_id = result["data"].get("session_id")

follow_up = requests.post(
    "https://api.lontiq.com/api/v1/chat",
    headers={
        "X-API-Key": "lntq_your_api_key_here",
        "Content-Type": "application/json"
    },
    json={
        "question": "What projects are they working on?",
        "session_id": session_id
    }
)

{
  "status": "Success",
  "data": {
    "answer": "Based on your data, the top performers in engineering are...",
    "sources": [
      "performance_reviews_q4.xlsx",
      "team_structure.csv"
    ],
    "audit": {
      "tool_calls": [
        { "tool": "get_schema", "duration_ms": 45 },
        { "tool": "search_nodes", "query": "top performers engineering" },
        { "tool": "run_cypher", "query": "MATCH (e:Employee)..." }
      ]
    },
    "graph_data": {
      "nodes": [...],
      "edges": [...]
    }
  },
  "message": null
}

MCP — Claude Desktop & Claude Code

Lontiq exposes a Model Context Protocol (MCP) server, allowing Claude to query your knowledge graph directly from Claude Desktop or Claude Code. Claude gets access to four tools:

Setup

Add the Lontiq MCP server to your Claude Desktop configuration file (claude_desktop_config.json):

{
  "mcpServers": {
    "lontiq": {
      "command": "python",
      "args": ["-m", "app.mcp"],
      "env": {
        "LONTIQ_API_KEY": "lntq_your_api_key_here"
      }
    }
  }
}

Once configured, Claude can autonomously query your knowledge graph, search for entities, and traverse relationships to answer questions about your data.

Status Codes

All error responses include a message field with a human-readable explanation:

{
  "status": "Error",
  "data": null,
  "message": "Write operations are not allowed. Use read-only Cypher queries."
}

Rate Limits

API key requests are rate-limited to 100 requests per minute using a sliding window. If you exceed this limit, you will receive a 429 response. Wait for the window to reset before retrying.

Permissions

Each API key is granted specific permissions that control which endpoints it can access. Permissions are assigned at key creation time by an admin.