AI & LLM APIs
Ask this question out loud to Siri and you will get a polite non-answer. Ask it in a developer forum and you will get a much longer, more interesting conversation. Can Siri AI call APIs? The short answer is yes, but not in the direct way most people imagine. Siri…
Ask a few iOS developers what “building an AI app” means right now and you will get a few different answers. Some mean wiring up Apple’s on-device Foundation Models framework. Others mean calling out to a cloud model provider over a REST API. Most end up doing a bit of…
Every few years, Apple ships an update that quietly changes what developers are expected to build. iOS 27 is one of those updates. Siri AI, the rebuilt assistant powered by the latest generation of Apple Intelligence, is not just a smarter voice assistant bolted onto the old framework. It is…
A couple of years ago, an iOS developer’s API surface area was fairly predictable. You had your backend REST endpoints, maybe a GraphQL layer, and a handful of third party services for payments or analytics. That world has gotten a lot busier. Model providers ship new endpoints every few months,…
Not long ago, the idea of doing real API development work on an iPhone would have gotten a laugh in most engineering standups. Phones and tablets were for checking Slack between meetings, not for testing endpoints, debugging authentication flows, or wiring up AI tools. That assumption is starting to look…
Every developer who has integrated an AI model into a product knows the moment. You have your API key, the documentation is open in another tab, and you are about to fire off your first request wondering if it will actually work or throw back a cryptic error. Testing Anthropic…
A few years ago, nobody’s API toolkit needed to think about REST, GraphQL, and WebSockets all at once, and then it did, because that is just what modern development looks like now. The same shift is happening again, faster this time, and it is called MCP. If your toolkit only…
You have swapped in the new model, updated the endpoint, and now you just need to know one thing: does the request actually return what you expect. Not eventually, not after you dig out your laptop, but right now, from wherever you happen to be. That is the entire premise…
A year ago, testing an AI endpoint meant pasting a cURL command into a terminal and parsing the response manually. Today, developers are hitting OpenAI, Anthropic, Google Gemini, Mistral, and a growing list of open-source model APIs dozens of times a day. They are tweaking system prompts, validating structured outputs,…
Not long ago, testing an API was a relatively predictable exercise. You sent a request, you got a response, and you checked whether the response matched what the documentation said it should be. Pass or fail. The logic was deterministic, the outputs were consistent, and your test suite could be…
