Browse categories
Explore
Fiverr Pro
English
$
USD
I will integrate chatgpt, gemini or claude into your ios app
Yevhenii P
About this gig
Hi, I'm Yevhenii an iOS developer with AI-powered apps live on the App Store. I connect Swift apps to language models and make the result feel fast, even when the model is slow.
WHAT I BUILD
- Streaming responses, so text appears as it generates instead of after a long wait
- Chat interfaces with conversation history and context management
- Vision features send photos or camera input to the model
- Speech in and out, using AVFoundation and the provider's audio APIs
- Structured output parsed into real Swift types, not raw strings
- Usage limits and cost controls, so one user can't drain your budget
PROVIDERS
OpenAI, Anthropic, Google Gemini, DeepL, or your own hosted endpoint. I can also set up fallback between providers, so one outage doesn't take your feature down.
ABOUT API KEYS
Shipping a key inside the app means it will be extracted. I'll tell you honestly whether you need a backend proxy for your case, and set one up if you do.
WHAT'S NOT INCLUDED
I don't train or fine-tune models, and I don't cover your API usage costs those stay on your own provider account.
Tell me what the feature should do and I'll tell you which package fits.
Get to know Yevhenii P
Yevhenii P
iOS Developer
- FromUkraine
- Member sinceAug 2026
- Avg. response time1 hour
Languages
Ukrainian, Russian, English
I'm an iOS developer from Ukraine, 5+ years with Swift. I ship native apps end to end: SwiftUI and UIKIt interfaces, clean MVVM code, in-app purchases with StoreKit 2, API and AI integrations, and App Store submission.
I've published my own apps, so I've been on both sides — writing the code and dealing with review rejections, crash reports and paywall conversion.
What you get from me: clear communication in English, no disappearing mid-project, and an honest "this won't work" when that's the truth. Message me before ordering and I'll tell you exactly what your project needs.
My Portfolio
FAQ
Can I just put the API key in the app?
You can, and people do — but keys inside an app binary get extracted, and then someone else spends your money. For a personal project it may be an acceptable risk. For anything published, I'll recommend a proxy and explain what it costs to run.
Who pays for the API usage?
You do, on your own provider account. I'll help you estimate what your usage will cost before we start, so there are no surprises.
Which model should I use?
Depends on the task, not on the hype. Tell me what the feature does and I'll suggest one, including whether a cheaper model would be good enough.
Can you make the responses faster?
Streaming makes it feel much faster, which usually matters more than raw speed. Beyond that, caching and prompt size are the main levers.
Do you fine-tune models?
No. If your task genuinely needs fine-tuning I'll tell you, but most cases are better solved with a better prompt and structured output.
