VIN Decoder MCP Server
Connect any AI assistant to live U.S. vehicle data. Decode any 17-character VIN and look up open NHTSA recalls right inside the chat — Claude Desktop, Claude Code, Cursor, ChatGPT, Continue, or any custom MCP client. Free, no auth, copy-paste install.
https://try.vin/mcp/vin- Transport
- HTTP (Streamable)
- Authentication
- None — public, rate-limited
- Tools
- 2 read-only lookups
Install in your AI client
Pick your client and copy the snippet. Restart the client and the decode_vin and lookup_recalls tools will appear in the tool picker.
Claude Desktop
Open Claude Desktop → Settings → Developer → Edit Config. Merge into the JSON, save, and restart.
{
"mcpServers": {
"try-vin": {
"type": "http",
"url": "https://try.vin/mcp/vin"
}
}
}Claude Code
Run this in your terminal. Add --scope user for global install. Verify with /mcp.
claude mcp add --transport http try-vin https://try.vin/mcp/vinCursor
Open Settings → Cursor Settings → MCP and add this entry. Restart Cursor.
{
"mcpServers": {
"try-vin": {
"url": "https://try.vin/mcp/vin"
}
}
}ChatGPT
ChatGPT supports MCP via custom connectors. Add a new connector with these values.
Settings → Connectors → Add a custom MCP connector
Name: try-vin
URL: https://try.vin/mcp/vin
Auth: NoneAnthropic SDK
Use the MCP server from any custom integration via the mcp_servers parameter on messages.create.
import Anthropic from '@anthropic-ai/sdk';
const client = new Anthropic();
const message = await client.messages.create({
model: 'claude-opus-4-7',
max_tokens: 1024,
messages: [
{ role: 'user', content: 'Decode VIN 1HGCM82633A123456 and check for recalls.' },
],
mcp_servers: [
{
type: 'url',
url: 'https://try.vin/mcp/vin',
name: 'try-vin',
},
],
});Tools
These are the tools the MCP server advertises to your AI client. Each one has a JSON schema the AI uses to call it correctly without any glue code on your side.
decode_vinDecode VIN
Decode a 17-character Vehicle Identification Number into structured vehicle data — year, make, model, trim, series, engine displacement, engine model, and country of assembly. Sourced from NHTSA's vPIC dataset.
| Argument | Type | Required | Description |
|---|---|---|---|
vin | string | yes | The 17-character VIN. Case-insensitive. Cannot contain the letters I, O, or Q. |
lookup_recallsLook up NHTSA recalls
Look up open NHTSA safety recalls for a specific U.S. vehicle by make, model, and model year. Returns each recall campaign number, summary, consequence, remedy, and the official NHTSA URL.
| Argument | Type | Required | Description |
|---|---|---|---|
make | string | yes | Vehicle manufacturer (e.g. "Honda", "Ford", "Toyota"). |
model | string | yes | Vehicle model (e.g. "Accord", "F-150", "Camry"). |
year | integer | yes | Four-digit model year, 1981 or later. |
Try a prompt
Once the MCP server is installed, paste any of these into your AI client. The model will automatically pick the right tool and call it.
Decode a VIN from a listing
"I'm looking at a used car listing with VIN 1HGCM82633A123456. Decode it and tell me what year, make, model, and engine I'm about to look at."
Check recalls before a test drive
"I'm about to test-drive a 2018 Honda Accord. Look up any open NHTSA recalls so I can mention them to the seller."
One-shot used-car vetting
"Here's a VIN: 5YJ3E1EA7JF005893. Decode it, then look up every NHTSA recall on file for that year/make/model. Summarize anything I should park-it-now or worry about."
Triage a fleet of VINs
"I have a fleet of 5 VINs (1FTFW1ET5DKE82010, 1HGCM82633A123456, 5YJ3E1EA7JF005893, JTDKB20U887763811, 4T1B11HK5JU123456). Decode each one and group them by make."