Authenticating

This page describes how to authenticate with various Smarty APIs. All requests must be authenticated.

There are two ways to authenticate your requests:

• embedded keys For NON server-side code (typically Android or iOS mobile devices, Javascript-powered web sites, and/or desktop applications).
• secret keys Only for code running in a trusted environment.

To view your keys, and to generate new ones, log in to your account and go to the API keys page.

Embedded keys

Embedded key are associated with one or more hostnames and/or IP addresses which you specify. Each of the hostnames (indicated by the 'Referer' header or 'Origin' header) will be granted rate-limited permission to call the API with the associated embedded key. For example, to use Smarty on your website where yoursite.com appears in the address bar, an embedded key like 3550738597428721952 must have the hostname yoursite.com associated with it. Each Embedded Key may have a maximun of 100 hosts.

Each IP address associated with an embedded key is granted un-throttled access to the API with the associated embedded key.

Note that the host yoursite.com is different from www.yoursite.com and both must be associated with that embedded key in order for it to work on both hosts.

We also support wildcard subdomains; e.g., *.yoursite.com will work. (Note that *.yoursite.com would not include yoursite.com, so you would need to use both in this case.)

If you are working in localhost, you can simply add "localhost" (without the port number) as a host paired with your embedded key. If you do so, remember to delete "localhost" as a host as soon as possible, as it makes your embedded key more vulnerable to unauthorized use.

Using embedded keys

To authenticate API requests with an embedded key, include key= and the embedded key in the query string. For example (line breaks have been added for readability):

https://us-street.api.smarty.com/street-address
	?street=123+main+Schenectady+NY
	&key=4236410529599436

Example use cases: Embedded key with hostname

• A website which calls any of our APIs using client-side Javascript (jQuery, angular, etc.).
• An iOS or Android application (phone/tablet) that users download and install from an 'app store'. (These apps should be coded to set the HTTP 'Referer' header (or 'Origin' header) to a hostname/IP address listed with the embedded key.)
• A downloadable/installable application meant to be used in an untrusted desktop environment. (These apps should be coded to set the HTTP 'Referer' header (or 'Origin' header) to a hostname/IP address listed with the embedded key.)

Here's a good example of setting the referer in a cURL request:

curl -v "https://us-zipcode.api.smarty.com/lookup?key=3550738597428721952&city=mountain+view&state=CA&zipcode=94035" --referer https://10.212.22.45

Example use cases: Embedded key with IP address

• A website which calls any of our APIs using client-side Javascript which is used exclusively by agents in a call center of your organization. In this scenario, it is likely that usage of the API is frequent enough that it would be rate-limited if a hostname was used. Associating an IP address with your embedded key removes the rate limit. Please note that the plan you subscribe to can also affect your rate limits regardless of your authentication method.

Rate limiting

When using public embedded key authentication (and depending upon the plan you are subscribed to), we restrict the number of requests coming from a given source over too short of a period of time. If you use embedded key authentication, you can avoid rate limiting by adding your IP address as an authorized host for the embedded key in question. This is done in order to prevent runaway charges caused by such conditions as a misbehaving (infinite) loop sending the same record over and over to the API. You're welcome.

Secret keys

Secret keys must be kept secret (hence the name). This means they should not be embededded or exposed in HTML or Javascript of public-facing web applications or untrusted desktop and mobile applications. Use secret keys only in code executing in a trusted or controlled environment that connects directly to the various Smarty APIs. A secret key is not limited to any particular hostname like embedded keys are. Secret keys consist of pair of values—an ID to identify your account along with a corresponding token which is like a password. While the ID value is safe to disclose over insecure or plaintext channels (such as emails to our customer service team), the token portion must remain secure and private at all times. WARNING: Disclosing, publishing, or broadcasting of secret key pair values is a direct violation of the Terms of Service and may be grounds for immediate termation of account privileges without prior notice or warning.

Using secret keys

To authenticate API requests with a secret key pair, specify auth-id and auth-token, containing the URL-encoded ID and associated token, respectively. For example (line breaks have been added for readability):

https://us-street.api.smarty.com/street-address
	?street=123+main+Schenectady+NY
	&auth-id=8d497be5-e211-4949-a18f-0bfd1d9970d3
	&auth-token=th4hargQiuyG7w7L7xfO

Example Use Cases: Secret Keys

• A server-side web application which calls any of the Smarty APIs from server-side code (PHP, .NET, Java, Python, Ruby, etc.).
• A server-side web application which calls any of the Smarty APIs from Javascript code (Node.js).