Anuma Launches Private AI Platform with Cross-Model Memory

The AI landscape just got more fragmented — or at least, it should have. Anuma launched publicly with a counterintuitive promise: one subscription unlocks ChatGPT, Claude, Gemini, Grok, DeepSeek, and other leading models, while keeping your conversation history encrypted on your device. The product emerged from a beta that attracted more than 10,000 users before opening to the general public.

What actually makes this different from another AI aggregator? The memory architecture. Most AI platforms trap your context inside their own ecosystem. Switch models, and you start from scratch. Anuma's encrypted vault stores preferences, projects, writing style, goals, files, and conversation history locally. That memory loads automatically into conversations across models, so you don't have to repeat yourself or rebuild context every time you switch tools (a problem that has plagued users for years, frankly).

According to the ZetaChain blog post from March 2026, Anuma is the first consumer app built on ZetaChain 2.0, a universal layer for AI and Web3. The platform enables applications to operate across multiple AI models while preserving private user context. Every new Anuma user creates a wallet on ZetaChain — not a cryptocurrency wallet in the traditional sense, but a cryptographic identity that serves as the user's encryption key and authentication layer.

The technical implementation matters here. When users import their ChatGPT, Claude, or Grok 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. Users can review, edit, delete, import, or export their vault whenever they want.

Core Contributor Ankur Nandwani previously co-created Basic Attention Token, the token that powers Brave's privacy-first browser economy. Brave's transparency reporting shows 112.2 million monthly active users as of March 31, 2026. That scale matters because it proves that privacy is not a niche preference when the product is strong enough. Nandwani's statement to the press captures the philosophy: "Today's AI products want to trap your context inside their own ecosystem. We built Anuma so your memory stays yours, no matter which model you use."

The physical experience of using Anuma differs from typical AI interfaces. Instead of navigating between separate apps or browser tabs, users switch between leading models mid-conversation without losing context or maintaining separate subscriptions. Council Mode compares answers from several models side by side. Users can also text AI over SMS and iMessage with no app download needed — the interface appears in your existing messaging app, not a new one.

Pricing sits at a free tier with 100 credits per month, Starter at $9.99 per month, and Pro at $19.99 per month. No credit card is required for the free tier. The credit-based payment system operates on ZetaChain through an escrow-based billing model, decoupling payment processing from centralized payment providers. Users purchase credits through subscriptions or credit packs that are consumed fractionally based on actual compute costs per model.

Independent reporting from Chainwire in January 2026 documented the beta launch and public waitlist. The article noted that McKinsey found ChatGPT reached 100 million users in two months, and OpenAI reported 800 million weekly active users by late 2025. Yet only 9% of consumers pay for more than one AI subscription across major assistants. This combination creates lock-in at the model layer and forces developers to repeatedly rebuild the same integration, routing, state, and billing infrastructure.

What Anuma actually solves is the friction of context loss. Imagine explaining your project to ChatGPT, then switching to Claude for a different perspective, then to Gemini for image generation. Without Anuma, you'd need to re-explain everything each time. With Anuma, the encrypted memory vault carries your context across all three models. The physical reality: fewer clicks, less repetition, no copy-pasting your own conversation history between apps.

The platform is available now on the web, with iOS and Android coming soon. Future features include building working apps from a prompt or description and solving real-world tasks such as billing, paperwork, and housing issues with Anuma agents. 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.

Whether users actually pay for this remains the real question. The free tier includes every major feature, which is unusual for AI platforms that typically gate advanced capabilities behind paywalls. But the market has shown that convenience often beats privacy in practice. Anuma's bet is that users will value owning their memory enough to pay for it — or at least stick with it long enough to prove the model works.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn