Anuma

Build AI-powered apps with a unified API, persistent memory, and user-owned encryption.

Building AI applications typically means managing multiple provider SDKs, handling API key logistics, and figuring out how to persist conversations. Anuma simplifies this by giving you access to models from OpenAI, Google, Anthropic, xAI, and open-source providers through a single API. Switch models without changing code, store semantic memories that persist across sessions, and keep user data encrypted with keys only they control.

Your users pay for their own inference directly — no API cost subsidies or billing management needed.

Quick Start


npm install @reverbia/sdk

The simplest way to send a message:


import { useChat } from "@reverbia/sdk/react";
 
const { sendMessage, isLoading } = useChat({
  getToken: async () => authToken,
  onData: (chunk) => setResponse((prev) => prev + chunk),
});
 
await sendMessage({
  messages: [{ role: "user", content: [{ type: "text", text: "Hello!" }] }],
  model: "gpt-4o-mini",
});

For persistent conversations with message history, use useChatStorage which adds automatic WatermelonDB storage on top of useChat.

Authentication Setup → | Full Tutorial →

What’s Included

The SDK provides chat completions with streaming and tool calling, persistent memory that searches past conversations semantically across sessions, server and client-side tools for web search, calendars, and custom functions, and automatic file processing for PDFs, Word docs, Excel, and images with OCR.

For the full API surface, see the SDK Reference.

Examples

Next.js React Native

Portal API

For other languages or direct HTTP access. The API is OpenAI-compatible, so you can use existing tools and libraries.

API Reference →