First App on ZetaChain 2.0 Lets You Import and Encrypt Your Entire AI History

Anuma’s Memory Import is the first real-world demonstration of ZetaChain 2.0’s private memory and AI interoperability infrastructure enabling cross-platform AI data portability secured by on-chain encryption.

Over 900 million people use AI chat weekly, but switching means starting from scratch. Your AI memory and months of conversations, preferences, context stays locked on corporate servers, often used to train the next model.

Today, Anuma launches Memory Import: export your full ChatGPT, Claude, or Grok history and bring it into Anuma’s encrypted Memory Vault in under 60 seconds. Everything is encrypted client-side before it reaches any server. Anuma is the first consumer app built on ZetaChain 2.0, the Universal Layer for AI and Web3, and the first to make AI memory portable, private, and user-owned at the protocol level.

Watch the demo video.

How Anuma integrates ZetaChain

ZetaChain is the Universal Layer for AI and Web3 that enables applications to operate across multiple AI models and blockchain networks while preserving private user context. Anuma is the first consumer-facing showcase of this architecture.

Identity and Wallet Creation

Every new Anuma user creates a wallet on ZetaChain. This isn’t a cryptocurrency wallet in the traditional sense. Rather, it’s a cryptographic identity that serves as the user’s encryption key and authentication layer. Users can sign up with Gmail or existing credentials and ZetaChain handles key generation and management transparently in the background. The wallet gives users sole ownership of the keys that encrypt their AI memory.

Private Memory Layer

ZetaChain provides a protocol-level memory system that is encrypted, permissioned, and persistent across sessions and AI models. When a user imports their ChatGPT history into Anuma, the data is encrypted client-side using AES-GCM with ZetaChain-managed keys before being stored in the Memory Vault. This is encryption by default. Not even Anuma’s servers can read the data.

AI Portal and Multi-Model Routing

ZetaChain’s AI Portal provides a unified routing and execution layer across multiple AI models, including GPT, Claude, Gemini, Grok, DeepSeek, and others. The portal handles availability, fallback, and cost-performance optimization automatically. Users interact with one interface and selects the preferred model for each query. Critically, the user’s memory persists across model switches so context is never lost when changing providers.

Payment Infrastructure

Anuma’s credit-based payment system operates on ZetaChain through an escrow-based billing model. This decouples payment processing from centralized payment providers and enables transparent, verifiable transactions that settle via stablecoins on ZetaChain. Users purchase credits through subscriptions or credit packs that are consumed fractionally based on actual compute costs per model.

Roadmap: Inference and Memory Decentralization

Looking ahead, ZetaChain and Anuma plan to progressively decentralize both inference execution and memory storage. This means AI computation will be distributed across a decentralized network rather than concentrated on centralized servers, and user memory vaults will be verifiably stored and encrypted on-chain. This roadmap extends ZetaChain 2.0’s vision of a universal layer where AI operates with the same self-custodial and portability principles that Web3 brought to financial assets.The private chat app Anuma is the first consumer proof point, but the developer SDK will soon be available for others to build on the same foundation.

About ZetaChain

ZetaChain is the universal layer for AI and Web3, letting developers build apps that run across chains and models, keep memory private, and monetize without infrastructure. With native connectivity across major blockchains and an AI interoperability stack powered by a Private Memory Layer, ZetaChain is building the foundation for the next generation of apps, agents, and experiences.

About Anuma

Anuma is a private AI memory app that lets you chat with any leading AI model including ChatGPT, Claude, Gemini, Grok, DeepSeek, and more, through a single encrypted interface. Unlike single-model apps, Anuma gives every user a personal Private Memory Layer that persists context across models, sessions, and devices. Memories are encrypted with AES-GCM, stored locally, and never used to train external models.

Categories