A living EV truth layer for AI agents.

$ curl https://www.evspec.io/v1/agent/resolve?make=Tesla&model=Model+3&version=RWD&year=2024

{
  "resolution": "unique",
  "match_quality": "exact",
  "should_clarify": false,
  "matches": [{
    "vehicle_id": "44125cfc-...",
    "make": "Tesla",
    "model": "Model 3",
    "version": "RWD",
    "model_year": 2024,
    "market": "UK",
    "match_reason": "exact_make_model_version_year"
  }]
}

$ curl https://www.evspec.io/v1/agent/vehicles/44125cfc-.../truth

{
  "identity": { /* make, model, version, year, market */ },
  "fields": [{
    "field_name": "range_wltp_km",
    "value": 513,
    "unit": "km",
    "status": "concordant",
    "proof_status": "verified",
    "primary_source_name": "OEM Official",
    "last_recorded_at": "2026-04-19T..."
  }]
}

# Install once (Claude Desktop, Cursor, any MCP client)
npx evspec-mcp

# Tools exposed to the agent
resolve_vehicle({make, model, year, version})
get_vehicle_truth({vehicle_id})
get_vehicle_changes({vehicle_id, since, limit})
explain_vehicle_field({vehicle_id, field_name})

# Add EV expertise to your agent
GET https://www.evspec.io/v1/agent-context

# MCP config (Claude Desktop / Cursor / any MCP host)
{
  "evspec": {
    "command": "npx",
    "args": ["evspec-mcp"]
  }
}

# Free tier — no key needed (20 req/min)
$ curl https://www.evspec.io/v1/agent/resolve?make=Tesla

# Developer tier — with API key (100 req/min)
$ curl https://www.evspec.io/v1/agent/resolve?make=Tesla \
  -H "X-API-Key: evspec_live_..."

# Request a key
contact@evspec.io