Integraph

API & MCP

Access all of Integraph’s structural engineering tools programmatically. Build automations, integrate with CI/CD pipelines, or let AI assistants run design calculations for you.

What is an API?

An API (Application Programming Interface) lets software talk to other software. Instead of using the web interface manually, you can write scripts or use other tools to create projects, run analyses, and generate reports automatically.

Integraph provides a REST API — the same standard used by services like GitHub, Stripe, and Google Maps. If you’ve used any web API before, ours works the same way.

View API Reference

What is MCP?

MCP (Model Context Protocol) is an open standard that enables AI assistants to interact with external tools. When you connect an AI assistant to Integraph via MCP, it can create sections, run analyses, and generate reports — just by asking in natural language.

Think of it as giving your AI assistant direct access to the same engineering tools you use in the browser. It can explore designs, run parametric studies, and automate repetitive tasks.

Getting Started

API Endpoints

All endpoints are available under /api/ on the app domain. Full documentation with request/response schemas is available via the interactive API reference.

Base URL: https://app.integraph.com.au/api

Docs: https://app.integraph.com.au/scalar/v1

MCP Configuration

Add the following to your AI assistant's MCP configuration to connect to Integraph. MCP is supported by Claude, ChatGPT, Gemini, Cursor, and other AI tools. 

{
  "mcpServers": {
    "integraph": {
      "type": "http",
      "url": "https://app.integraph.com.au/mcp"
    }
  }
}

The MCP endpoint exposes project management, section analysis, LGS design, concrete design, and report generation as AI-callable tools.

Predictable, per-call pricing

Pro-tier API and MCP calls are priced per endpoint in AUD. Every call advertises its cost up front — in the OpenAPI document, on the MCP tool description, and on the response itself — so scripts and AI agents can reason about spend before they make a request.

Discovery endpoint

GET /api/billing/api-costs returns the full endpoint → cost map, the AUD currency, and the platform markup. Fetch it once at startup; cache as long as you like.

View pricing catalogue Raw JSON (for scripts & agents)

Response headers

Every metered response carries:

  • X-Api-Cost-Micro-Dollars — what this call cost
  • X-Api-Balance-Micro-Dollars — balance remaining
  • X-Api-Period-End — when the next refill lands

OpenAPI & MCP

Metered REST endpoints carry a cost extension visible in the interactive API reference. Every MCP tool description ends with (Cost: $X.XXXXXX AUD per call) so AI agents pick it up at tool-discovery time.

Hard stop on exhaustion

When the balance hits zero, the next metered call returns 402 Payment Required with code: api_balance_exhausted. No overage, no surprise invoice — you choose when to top up.

See Subscriptions and trials for the included allowance, refill cadence, and top-up workflow.

Use Cases

Parametric Studies

Use the API or MCP to run hundreds of design iterations and compare results automatically.

CI/CD Integration

Run structural checks as part of your build pipeline. Validate designs on every commit.

AI-Assisted Design

Let AI assistants explore design alternatives, optimise sections, and generate reports conversationally. Works with Claude, ChatGPT, Gemini, Cursor, and more.

Data Migration

Import sections and materials from existing systems via the REST API.

Custom Dashboards

Build internal tools that pull live section properties and design results.

Automated Reports

Generate PDF reports programmatically for large project batches.

Ready to modernise your design workflow?

Join engineers who are designing smarter, not harder.

Get Started Free

No credit card required. Free tier available forever.