Docs

Why use AIProxy?

API keys cannot be kept secret inside a shipped app. Loading a key at runtime does not solve the problem: to use the key over the network, the app must receive the complete credential and attach it to a request, where someone who controls the client can extract and reuse it. AIProxy lets your app call AI providers without ever giving the client the full provider key, while adding controls and visibility that are difficult to build and operate yourself.

Protect provider credentials: Your complete API key is never included in the app binary or sent by the client.
Prevent endpoint abuse: Use Apple DeviceCheck, endpoint restrictions, and per-user or per-IP rate limits to control who can make requests and how often.
Monitor production usage: Inspect request history, errors, token usage, client activity, and traffic trends in the dashboard.
Integrate quickly: Use our open-source Swift package for typed access to popular AI services, or proxy requests from your existing networking code.

How AIProxy works

To use AIProxy you first add the service/API you want to protect, then add your API key through our dashboard. We don't actually store the key on our servers, we encrypt the key and store 1/2 of that result in our backend and give the customer the other 1/2 to send up with requests to AIProxy. We marry the two pieces and decrypt the key, forwarding that onto the API. Learn more

View integration guide Sample apps Swift client library

Supported Services

Service	Supported	Endpoints	Examples
Anthropic	Yes	Messages, streaming-messages, streaming tool calls, send image, tools API	Examples
Anthropic Migration (0.135.0)	Yes	Migration guide for Anthropic SDK	Examples
Azure	Yes	OpenAI	Examples
Brave Search	Yes	Search	Examples
DeepL	Yes	Translations	Examples
DeepSeek	Yes	Chat completion, streaming chat completion	Examples
EachAI	Yes	Each AI Workflows	Examples
ElevenLabs	Yes	Text-to-speech	Examples
Fal	Yes	FastSDXL, Runway Gen3 Alpha, Flux training	Examples
Fireworks AI	Yes	DeepSeek R1	Examples
Gemini	Yes	Text generation, audio transcription, vision	Examples
Groq	Yes	Non-streaming chat, streaming chat, transcribe audio	Examples
Mistral	Yes	Chat, streaming chat	Examples
OpenAI	Yes	Chat, streaming chat, multi-modal chat, DALLE, text-to-speech	Examples
OpenRouter	Yes	Chat completions, streaming chat completions, structured outputs (strict json), multi-modal (image input)	Examples
Open-Meteo	Yes	Fetch the weather	Examples
Perplexity	Yes	Chat completion, streaming chat completion	Examples
Replicate	Yes	Flux-Schnell, Flux-Dev, Flux-Pro, Flux-PuLID, Flux ControlNet, SDXL, Flux fine-tune	Examples
Stability.ai	Yes	Image generation	Examples
Together AI	Yes	Chat, streaming chat, tool call with Llama	Examples