AI Integration

Use MotiveGrid inside ChatGPT & Claude

Connect our free, read-only MCP server (Model Context Protocol) and your AI assistant can compare cars, pull five-year cost of ownership, check safety and driver-assistance ratings, and recommend the right vehicle — every answer grounded in real data, never made up.

Server endpoint

https://motivegrid.com/api/mcp

How to connect →Try it on the site

What you can ask

Once connected, just ask in plain language.

“Compare the 2026 Toyota RAV4 and Honda CR-V on five-year cost of ownership.”

“What’s the best compact SUV for a tall driver who wants low running costs?”

“Show me the MotiveGrid scorecard for the 2026 Hyundai Ioniq 5.”

“I have $40k, need 3 rows and great safety — what should I look at?”

Capabilities

Four grounded tools your assistant gains.

Find a vehicle

Resolve any model we track to its data — “the 2026 RAV4”, “electric SUVs under $50k”, “Honda’s compact crossover”.

Full scorecard

Pull a vehicle’s MotiveGrid Score, class percentile, and a plain-language read on cost, powertrain, safety, driver assistance, and livability.

Head-to-head compare

Two cars side by side: itemized five-year cost of ownership, key specs, and a grounded verdict on the better long-term pick.

Recommend the right car

Describe a buyer — budget, size, priorities — and get a ranked shortlist with the reasoning behind each pick.

Setup · 2 minutes

How to connect

Claude (web — easiest)

Open Settings → Connectors in Claude.
Click Add custom connector.
Paste the endpoint URL below and save. No extra software needed.

https://motivegrid.com/api/mcp

Claude Desktop

Open Settings → Developer → Edit Config, add the block below to claude_desktop_config.json, then fully quit Claude (⌘Q) and reopen.

{
  "mcpServers": {
    "motivegrid": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://motivegrid.com/api/mcp"]
    }
  }
}

macOS not finding it? Finder-launched apps get a minimal PATH that misses Node, so "npx" fails silently. Use this variant instead — the absolute npx path (yours may differ; run which npx in a terminal) plus an explicit PATH:

{
  "mcpServers": {
    "motivegrid": {
      "command": "/opt/homebrew/bin/npx",
      "args": ["-y", "mcp-remote", "https://motivegrid.com/api/mcp"],
      "env": { "PATH": "/opt/homebrew/bin:/usr/local/bin:/usr/bin:/bin" }
    }
  }
}

Running mcp-remote in a terminal only tests the endpoint — it does not register the server. The config above is what registers it.

ChatGPT

Connector support in ChatGPT depends on your plan and the ongoing rollout, so it may not be available in every workspace yet. If it is:

Enable Developer mode / Connectors in ChatGPT settings.
Add a new connector and paste the endpoint URL.
Start a chat and ask it to compare or recommend a car.

https://motivegrid.com/api/mcp

Free & read-only

No cost, no account, no write access — it only looks things up.

Grounded, never invented

Real database values only; unknown models return suggestions, not made-up numbers.

Primary-sourced data

Depreciation from CarEdge, fuel from the EIA, crash ratings from NHTSA & IIHS.

For developers

Protocol & tools

A remote MCP server over Streamable HTTP. The endpoint speaks JSON-RPC 2.0 over POST (initialize, tools/list, tools/call). It is read-only, needs no auth, and is rate-limited per IP.

Opening the endpoint in a browser returns an error by design — it is a JSON-RPC POST endpoint, not a web page. Use an MCP client or a POST request.

search_vehiclesResolve free text to vehicle slugs.

query, limit?

get_vehicle_scorecardFull scorecard for one vehicle.

slug, trim?

compare_vehiclesHead-to-head + 5-yr cost + verdict.

slug_a, slug_b, trim_a?, trim_b?

recommend_vehicleRanked shortlist from buyer priorities.

priorities[], budget_min_monthly?, budget_max_monthly?, category?

Full input schemas are returned live by tools/list. Slugs are canonical (e.g. toyota-rav4-2026) — resolve free text with search_vehicles first.

FAQ

Common questions

What is the MotiveGrid MCP server?

It is a remote Model Context Protocol (MCP) server — a standard way for AI assistants like ChatGPT and Claude to call live tools. Connecting it lets the assistant pull MotiveGrid’s structured car data directly: cost of ownership, safety, driver assistance, livability, and head-to-head comparisons. It is free and read-only.

Is it free?

Yes. The MotiveGrid MCP server is free to connect and use. It is read-only — it can look up and compare vehicles, but it cannot change anything or access your data.

Does the AI make up car data?

No. Every answer is grounded in MotiveGrid’s database. If you name a vehicle we do not track, the tool says so and suggests valid alternatives instead of inventing numbers. Depreciation comes from CarEdge, fuel prices from the EIA, and crash ratings from NHTSA and IIHS.

Which assistants does it work with?

Any MCP-capable client. Today that includes Claude (web and desktop) and ChatGPT in developer/connector mode, plus tools like Cursor. The same endpoint works everywhere: https://motivegrid.com/api/mcp.

It is not showing up in Claude Desktop — what do I do?

On macOS, apps launched from Finder get a minimal PATH that can miss your Node install, so npx fails silently. Use the absolute path to npx (run “which npx” in a terminal) as the "command", add an "env" block with your PATH, then fully quit Claude (Cmd-Q) and reopen. Running mcp-remote in a terminal only tests the endpoint — it does not register the server.

Compare cars on the site →How we score