Sui
The base network that gives Cortex fast settlement, composable ownership, and identity-native coordination.
Cortex is the sovereign memory layer built on Sui, Walrus, and Seal. Keep durable encrypted memory, share it selectively, and turn it into prompts, loops, multi-agent workflows, and dreams.

Most memory tools stop at storage and search. They do not give you sovereign ownership, encrypted shared memory, or programmable workflows for prompts, loops, and agents. Your context lives in someone else's product instead of on infrastructure aligned with you.
Cortex works in three steps: store sovereign memory on Sui, coordinate shared and agentic workflows across it, and extract higher-order insight from everything you have already saved.

Cortex is built on Sui, Walrus, and Seal so your memory is durable, encrypted, and aligned with ownership instead of platform lock-in.
Create encrypted shared memory between people and agents. Cortex scopes access by identity so collaboration stays precise.
Turn grounded memory into better prompts instantly, so every model and tool starts from your actual context instead of a blank box.
Generate loops directly from memory, then let Cortex run iterative research, planning, and review cycles over the same knowledge base.
Specialist agents can interface over one memory layer, hand off work, and stay grounded in shared context rather than isolated chats.
Dreams synthesize patterns, weak signals, and long-horizon insight from memory so Cortex can surface what matters before you ask.
Cortex covers the same core memory jobs users expect, then goes further with an elastic brain that learns from usage, reduces memory bloat through consolidation sweeps, and layers Sovereignity, shared memory, prompts, loops, agents, and dreams on top.
Cortex | Supermemory | |
|---|---|---|
Memory recallPersistent memory search and retrieval across saved context | Best Fit | Yes |
Document and media ingestionBring in notes, files, links, audio, images, and video | Best Fit | Yes |
Shared memoryIdentity-scoped encrypted collaboration between people and agents | Native | No |
Prompt generationGenerate grounded prompts directly from saved memory | Native | No |
Loop generationTurn memory into agentic workflows and iterative runs | Native | No |
Multi-agent interfaceSpecialist agents work over one shared memory graph | Native | No |
DreamsMemory-derived insight and synthesis across your graph | Native | No |
Elastic brainLearns from use, runs consolidation sweeps, folds duplicates, and prunes redundant entries to reduce memory bloat over time | Native | No |
SovereignityBuilt on Sui with Walrus storage and Seal encryption | Yes | No |
This is the actual Cortex brain graph style, seeded with sample memories about the Cortex project. Drag it, zoom it, search it, open nodes, and inspect how the graph connects ideas inside an elastic memory system.
Cortex is built on the Sui ecosystem end to end, combining execution, storage, encryption, memory infrastructure, and naming into one sovereign memory stack.
The base network that gives Cortex fast settlement, composable ownership, and identity-native coordination.

Stores memory, files, and media as durable blobs so your context stays available beyond one device or one session.
Protects memory with programmable access control, so only the right users, agents, and workflows can decrypt it.
Keeps recall fast and agent-friendly with a memory layer designed for retrieval, updates, and durable context sync.

Turns addresses into readable names so sharing, discovery, and coordination feel native instead of wallet-shaped.
Get answers about how Cortex stores, shares, and activates sovereign memory across people, agents, and workflows. Your privacy, your infrastructure, your terms.
Launch Cortex and put your memory on infrastructure that can store, share, orchestrate, and dream with you. Built on Sui, designed for sovereign AI workflows.