For AI agents

IP intelligence without an auth flow.

A simple HTTP tool for assistants, automated checks and agent workflows.

Give your agent these instructions

Use IP Info for approximate public-IP geolocation and ASN lookup.
Authentication: none. Cost: free.
If an IP is supplied: GET https://ip-info.com/json?ip=<URL-encoded-IP>.
If checking your own execution environment: GET https://ip-info.com/json.
Do not describe your server/proxy exit as the human user's IP.
Use country for country-code comparisons and asn for numeric ASN checks.
Treat null as unavailable; never fabricate missing data.
Treat location and network data as approximate and potentially outdated.
Geolocation is approximate. This is not VPN/proxy detection.
Handle JSON errors and use bounded backoff for temporary failures.

Caller does not mean human user

When an agent calls /json, the response describes its runtime or proxy’s network exit. To check another person’s IP, supply an IP they have explicitly provided or your application has legitimately obtained. Do not infer it from the agent’s own connection.

Minimal integration

# Python 3; standard library only
import json
from urllib.request import urlopen

with urlopen("https://ip-info.com/json?ip=94.204.59.232", timeout=10) as response:
    print(json.load(response))

Predictable data, clear limitations

Responses contain explicit numeric, string, boolean and nullable fields. There are no login pages or API-key exchanges. Handle unavailable fields and preserve uncertainty around location and network classifications.

The human reference, OpenAPI schema and text reference are generated from the same response definition. No MCP server is required: ordinary HTTP is sufficient.