Last time I wrote about the difference between'open' vs. 'free' APIs.
Another interesting dimension is free vs. paid.
Free vs. Paid has to do with the business model of your service. If you have a service that’s valuable/revenue-generating to the user (unlike advertising/reach, which is valuable to *you*) then you might choose to ask to be paid - through a credit card, licensing or BD deal. (we've written before on the many API monetization models and resulting API roadmap and technology implications.)
While I hate the term "freemium" (more on that later) .. You might also choose to offer your free service up to some limit (100 API messages/hour) and then charge for larger volumes (100-999 calls cost X, 1000-4999 calls cost Y, etc.). The lowest rung on the pricing waterfall is often 'free'.
Twitter's API has both free and paid access. For typical end-users, the API is free and provides a limit of 150 messages per hour. For larger usages – such as those from media and news companies who want to harvest Twitter data – there is paid access to enable message volumes in the thousands and higher per hour.
No doubt if the volume of API calls in the Tweetdeck case to Facebook goes up, there could be a charge; or if Twitter starts selling licenses to consumers for larger API bandwidth as they do today for enterprises (media companies), Tweetdeck could be a place where those licenses are sold. Personally I would happily pay $2-5/month to get 500-1000 messages per hour on Twitter due to the incredible utility of Tweetdeck. Otherwise I’d be confined to the Twitter web interface which is just not that useful to me.
We are really in a very dynamic part of the API market right now where many business models are being tested out. We should have a longer conversation about the different business models that seem to be working. A good example - Fred Wilson had some great and different thinking that can open up new profit models through APIs.
Next - is there a business model in *paying* people to use your API?