In this Reddit post today, Christian Selig, the developer of the popular Reddit client Apollo, predicts his monthly API fees will soon be $20 million (US) per year.

A death sentence.

The new Twitter API rates are equally punishing. Even the lower-cost "API Pro" tier prices most of us out at 1M tweets per month for $5000 (US) per month. Ironically, many of the apps that could afford that tier shut down after the previous announcement of their plan with 50M tweets per month for $42,000 (US) per month, which they inexplicably announced first. (Also inexplicable: you can get 2x the tweets for $125,000 (US) per month, at a mere 3x the cost.)

Where does either company think good apps come from? Startups and independent developers are creating experiments to make a better experience in those ecosystems; they are rarely 'stealing' the customers, providing no value back to the platform. And, it's hard to predict how much new software will catch on: very few companies will lay out those kinds of investments before seeing customer traction. Either there's no funding for it, or it's a bad bet. A catch-22: no app -> no traction. No traction -> no way to justify such a high price tag to even enter the arena.

These apps bring customers to your platforms. They help people with getting more engagement, causing more platform usage. They make finding important and interesting things easier (like my product). They help people create new content, the fundamental value your entire businesses rest on.

At these rates, we might even cynically conclude these companies are trying to end third-party apps altogether. (With the paid plans in place as exceptions for the desparate because, hey, why throw out all revenue?).

Other possible reasons: fighting bots (including LLM training), being able to advertise more (and control what content is placed near the ads), or because it costs too much to run the APIs.

I see how any of these could be legitimate concerns if they are the true motives.

But I can not wrap my head around why these charges must be this exorbitant, or why they are predominantly tied to the application.

I've always had that issue with the Twitter APIs; it's not a new problem. Why is the tweet limit tied to all traffic generated from the application? A user delegates to a tool to act on their behalf, but it's not the user's limits that are in play.

It would be like charging browser developers for customer traffic, tallied up by the user-agent string. Or if AWS charged third party client libraries instead of the holder of the API key.

I realize things are trickier there when you have free sign-ups, but both platforms have subscription plans now. There are options that don't drive away your customers and all these developers who make your platform better.

Charge for it. Charge by usage. Charge the customer.

I'm not saying that's a better end-user experience than an open API. But, yes, maybe the the APIs are underwater and it's just not feasible.

If a customer wants to use a third-party tool with your service, you could charge them for it.

For example, say an introductory add-on for Twitter is $5/mo for 500,000 tweets per month, usable across all the tools they want to use. They could get 10,000 for free as a way to experiment with apps to see if they'll work out. The customer pays the app developers separately. Each of those vendors might pay $50/mo to be part of the ecosystem (as a way to offset support costs). You can make custom deals with "Big LLM" for billions of calls.

That will annoy many people who don't want to pay at all, but it’s better than shutting out third party apps entirely. And it could make you money as apps get popular. It could even be in your financial interest to promote them. And developers would be motivated to touch your platforms again.

An infrastructure company like AWS is, again, a model. People routinely pay third-party companies to create tools that call AWS and then pay AWS on a usage basis (because they've loaded their API keys into those tools). And AWS routinely helps their customers select good third-party tools.

You have these options. But instead, you're choosing to alienate a huge swath of developers and customers, ultimately hampering your own attempts at growth and revenue.