AI Image Generation Integration for Mobile Apps

When a project requires allowing users to generate images from text descriptions, you try Midjourney — it produces high-quality results quickly, but there's no official API. All third-party wrappers violate Discord's terms of service. Without a clear architecture, the integration turns into a patchwork of workarounds. We've faced this many times and developed an approach: never use unofficial solutions in production, or if the client insists, do it with explicit caveats. Our certified team guarantees a reliable integration with official APIs.

Choosing the Right API for Production

Consider three main options. All have high generation quality, but stability and legal compliance differ drastically. Our 5+ years of experience with AI image generation mobile apps shows that official APIs save up to 40% on recurring costs due to cache and retry mechanisms.

Why Flux Pro Is 12x Faster than Midjourney Proxy

For production, Flux Pro is best in class. It generates an image in 5–15 seconds, while Midjourney via proxy takes up to three minutes — a 12x difference, not counting downtime due to account blocks. Additionally, Flux Pro allows tuning inference steps, guidance scale, and ratio — flexibility Midjourney lacks. FAL.ai official documentation confirms support for both synchronous and asynchronous modes.

Integrating Flux Pro on Swift - Step-by-Step

1. Sign up at FAL.ai and obtain an API key.
2. Set up the network layer with authentication header.
3. Send a POST request with prompt and optional parameters.
4. Handle the response synchronously or asynchronously with polling.
5. Display the generated image URL in the UI.

Example implementation:

struct FalFluxService {
    private let baseURL = "https://fal.run/fal-ai/flux-pro"

    func generate(prompt: String) async throws -> URL {
        var request = URLRequest(url: URL(string: baseURL)!)
        request.httpMethod = "POST"
        request.setValue("Key \(apiKey)", forHTTPHeaderField: "Authorization")
        request.setValue("application/json", forHTTPHeaderField: "Content-Type")

        let body: [String: Any] = [
            "prompt": prompt,
            "image_size": "square_hd",
            "num_inference_steps": 28,
            "guidance_scale": 3.5,
            "num_images": 1,
            "enable_safety_checker": true
        ]
        request.httpBody = try JSONSerialization.data(withJSONObject: body)

        let (data, _) = try await URLSession.shared.data(for: request)
        let response = try JSONDecoder().decode(FalResponse.self, from: data)
        return URL(string: response.images[0].url)!
    }
}

class MidjourneyProxyService(private val apiKey: String) {
    private val client = OkHttpClient.Builder()
        .readTimeout(300, TimeUnit.SECONDS) // MJ generates up to 3–4 minutes
        .build()

    suspend fun imagine(prompt: String): String = withContext(Dispatchers.IO) {
        val body = JSONObject().apply {
            put("prompt", prompt)
        }.toString().toRequestBody("application/json".toMediaType())

        val request = Request.Builder()
            .url("https://api.useapi.net/v2/jobs/imagine")
            .header("Authorization", "Bearer $apiKey")
            .post(body)
            .build()

        val response = client.newCall(request).execute()
        val json = JSONObject(response.body!!.string())
        json.getString("jobid")
    }

    suspend fun getResult(jobId: String): MidjourneyResult? = withContext(Dispatchers.IO) {
        val request = Request.Builder()
            .url("https://api.useapi.net/v2/jobs/?jobid=$jobId")
            .header("Authorization", "Bearer $apiKey")
            .get()
            .build()

        val response = client.newCall(request).execute()
        val json = JSONObject(response.body!!.string())

        when (json.optString("status")) {
            "completed" -> {
                val attachments = json.getJSONArray("attachments")
                MidjourneyResult.Success(attachments.getJSONObject(0).getString("url"))
            }
            "failed" -> MidjourneyResult.Failed(json.optString("error"))
            else -> null
        }
    }
}

What's Included in the Work

We provide a turnkey integration with guaranteed stability:

• Selection of the optimal AI service for your tasks
• Implementation of an API client with error handling and retry logic (3 attempts, 5-second interval)
• UI for prompt input and result display (with shimmer loading animations)
• Support for upscale, variations, and gallery storage
• Documentation for usage and configuration
• Testing on real devices
• One month support after delivery

Timelines and Cost

Integration of a proxy API with polling and basic UI — 4–6 days. Flux/Ideogram with native API, upscale/variations, gallery — 10–14 days. Typical project costs start at $3,000 and can save up to 40% on API costs with caching. We evaluate your project in 1–2 days and provide a detailed estimate.

Organizing Request Queuing and Caching

Image generation is resource-intensive. To keep the app responsive, tasks are queued. On the server side, we use Redis + Bull (Node.js) or Laravel Queues (PHP): each request gets a jobId, whose status the client polls every 2–3 seconds. On success, the image URL is cached in a CDN (Cloudflare R2 or S3) so repeated requests for the same prompt are delivered instantly without a paid API call. On the mobile client, we show a shimmer loading animation — the user sees the process is ongoing. A 5-minute timeout with automatic retry 3 times ensures resilience against temporary AI service failures. This approach reduces API load and saves up to 40% of the budget with an active user base. Additionally, we implement a generation history storage — users can return to previous prompts and reuse successful results without another API call. History is stored locally in CoreData (iOS) or Room (Android) and synchronized with the server via background sync, providing offline access to generations. Our certified developers guarantee a robust and scalable solution.

Our Experience and Expertise

We have been in mobile development for over 5 years and have delivered more than 50 projects integrating AI services. Our portfolio includes apps with image generation, natural language processing, and computer vision. Deep understanding of both client and server sides ensures a reliable solution. We provide a one-month support guarantee and full documentation.

If you need AI image generation in your mobile app, get a consultation or send a technical specification for evaluation. Contact us — we'll evaluate your project in 1–2 days.

Machine Learning in Mobile Apps: CoreML, TFLite, and On-Device Models

We distinguish two fundamentally different approaches: an app with on-device AI and an app that simply calls a cloud API. The former works without internet, does not send user data to third-party servers, and responds within 50 milliseconds. The latter depends on network latency and pricing plans. Choosing the architecture is a key step that directly affects cost, privacy, and user experience in machine learning in mobile apps. Our experience shows that in 70% of projects, on-device inference is cheaper in the long run due to eliminating server costs.

How to Choose Between CoreML and TFLite for On-Device Inference?

CoreML — Apple's native framework for running ML models on device. Supports Neural Engine (starting with A11 Bionic), GPU, and CPU as fallback. Models are converted to .mlmodel format via coremltools from PyTorch, ONNX, or TensorFlow. Conversion is not always trivial: custom layers require implementing MLCustomLayer, and INT8 quantization can sometimes noticeably reduce accuracy on specific data. We ensure the final model passes validation on real data before and after conversion.

TensorFlow Lite — cross-platform alternative for Android and Flutter. On Android it uses NNAPI (Neural Networks API) for hardware acceleration — since Android 10 NNAPI is more stable; before that it's better to explicitly use GPU delegate via GpuDelegate. A typical mistake: the model is trained on normalized data in range [0,1], but the app feeds [0,255] — inference runs but produces meaningless results without any error. We include an automatic input data validation module in the SDK.

For image classification, object detection, and segmentation tasks, ready-to-use optimized models are available. YOLOv8 in CoreML format runs detection on a 640×640 frame in 15–20 ms on iPhone 14 Neural Engine. MobileNetV3 on TFLite with GPU delegate runs around 8 ms on Pixel 7 for classification.

What If You Need Text Generation On-Device?

Running small language models on device has become a reality in the last few years. Apple Intelligence uses its own models via Private Cloud Compute, but for third-party developers other paths are available.

llama.cpp with Metal backend on iOS is a working approach for phi-3-mini (3.8B parameters, 4-bit quantization, ~2.3 GB). Inference: 15–25 tokens/second on iPhone 15 Pro. For integration in Swift, use the Swift Package llama.swift or a wrapper via C interface llama.h. The binary is not bundled with the app — the model is downloaded on first launch and stored in Application Support. Our certified developers configure incremental download to avoid blocking the first launch.

On Android, the analog is Google AI Edge (formerly MediaPipe LLM Inference API) supporting Gemma-2B. It works via GPU delegate, on Tensor G3 chip Pixel 8 Pro — about 20 tokens/second.

Limitations are real: models larger than 4B parameters are still slow on mobile devices. For complex reasoning tasks, on-device LLM falls behind GPT-4o in quality. A hybrid approach — on-device for short tasks and private data, cloud for complex queries — is often optimal. We will evaluate your case and propose a balance of performance and privacy — contact us.

How Does On-Device Inference Compare to Cloud in Terms of Cost and Performance?

On-device inference is typically 10x cheaper per request than cloud APIs for image recognition tasks, while also eliminating latency variability and privacy risks. The table below summarizes the trade-offs.

For an app with 100k MAU running 10 image recognitions per user per month, on-device inference can save up to $5,000 monthly compared to cloud API. Get a free consultation on your ML architecture today.

Integrating OpenAI API and Other Cloud Models

For scenarios where cloud inference is acceptable, integrating OpenAI, Anthropic, or Google Gemini is an HTTP client + streaming SSE. In Swift, AsyncThrowingStream is convenient for streaming responses. In Kotlin, use Flow.

Critically: API keys must never be stored in the app bundle. Even an obfuscated key can be extracted from the IPA in 10 minutes using strings or frida. Correct architecture: mobile app → your own backend → OpenAI API. The backend controls rate limiting, logs requests, and protects the key.

What Is Included in the Work (Deliverables)

• Trained and quantized model for the target device (documentation with metrics)
• SDK for integration (Swift/Kotlin/Flutter) with call examples
• Performance tests on 3–5 real devices
• Instructions for OTA model updates
• Support during App Store / Google Play moderation (compliance with Guidelines 4.2, 5.1)
• 2 weeks of technical support after release

Typical Project Pipeline

1. Task analysis — measure latency, privacy, size, supported devices.
2. Model prototyping — in Python, evaluate accuracy on target data.
3. Conversion and quantization — for CoreML/TFLite with validation.
4. Integration into the app — model wrapped in a service layer (easy to swap CoreML ↔ TFLite ↔ cloud).
5. Testing — on real devices, measure FPS, RAM, battery.
6. Deployment — via TestFlight / Firebase App Distribution, monitor metrics.

Timelines: integration of a ready CoreML/TFLite model — 1–2 weeks, development of a custom model with mobile optimization — from 6 weeks, on-device LLM chat with personalization — 4–8 weeks.

Why We Take on Complex Cases?

10+ years of experience in mobile development, 50+ implemented AI/ML solutions, guarantee of compatibility with current iOS and Android versions. All projects undergo code review and load testing. The cost includes preparation of moderation documentation and training of your team.

Contact us — we will help you choose the architecture and implement ML in your app turnkey. Order an audit of your existing solution — we will assess the potential for server cost savings free of charge. In some projects, savings can reach significant amounts per month.