AI Integration

Reox makes Large Language Models a first-class citizen in your code.

The `ai` Keyword

Define AI assistants directly in your application code. Reox handles the context management and API calls to the system-level LLM service.

ai assistant {
    model: "neolyx-gpt"
    temperature: 0.7

    fn suggest_reply(email: string) -> string {
        return ai.prompt("Draft a polite reply to: {email}");
    }
}

Async AI Operations

AI operations can be performed asynchronously using async fn and await, keeping your UI responsive while waiting for model responses.

async fn generate_code(prompt: string) -> string {
    let result = await ai.prompt(prompt);
    return result;
}

// Usage in async context
async fn handle_button_click() {
    let code = await generate_code("Write a sorting function");
    display_result(code);
}

Native Performance

Unlike calling REST APIs in other languages, Reox AI calls are optimized IPC (Inter-Process Communication) messages to the NeolyxOS Neural Engine service, ensuring low latency and privacy.