Quickstart

This page builds the smallest useful flow: configure a model, generate one response, stream one response, and add request options. The same shape is used throughout the SDK.

You want	Start with
One complete answer	`generateText`
UI that updates while text arrives	`streamText`
Typed JSON	`generateObject`
Model-controlled app actions	`generateWithTools`
SwiftUI state without building async plumbing	`AIChat` or `AICompletion`

Create A Model

import SwiftyAI
 
let model = OpenAICompatibleProvider(
    baseURL: "https://api.openai.com/v1",
    apiKey: ProcessInfo.processInfo.environment["OPENAI_API_KEY"]!,
    model: "gpt-4o-mini"
)

Generate Text

Use generateText when your code needs the full answer before it can continue.

let answer = try await generateText(
    model: model,
    prompt: "What should a Swift developer know before using actors?",
    options: GenerationOptions(
        system: "Be practical and concise.",
        temperature: 0.3,
        maxTokens: 300
    )
)
 
print(answer.text)
print(answer.usage?.totalTokens ?? 0)
print(answer.finishReason ?? "unknown")

Stream Text

Use streamText when the UI should update as tokens arrive.

var transcript = ""
 
for try await chunk in streamText(
    model: model,
    prompt: "Write a short haiku about Swift concurrency."
) {
    transcript += chunk.text
    print(chunk.text, terminator: "")
}

The last chunk may include usage and finish reason if the provider sends that metadata.

Add A Stop Sequence

Options are passed per request. This keeps default model setup separate from request behavior.

let response = try await generateText(
    model: model,
    prompt: "Return three labels, then write END.",
    options: GenerationOptions(
        maxTokens: 120,
        stopSequences: ["END"],
        retryPolicy: RetryPolicy(maxAttempts: 3)
    )
)

Next Feature

Once the basic call works, move to generate object for typed JSON, tool calling for multi-step work, or SwiftUI hooks for app state.