Skip to main content
AI Memory Infrastructure Pricing

Persistent memory built for individuals & enterprises

Choose the tier that matches your scale. From individual developer experimentation to full organizational governance and self-hosted deployment.

Free

$0/forever

Experimentation & light context

  • ✓3 memory spaces
  • ✓200 memories
  • ✓Full-text & vector retrieval
  • ✓1 API key for AI connections
  • ✓MCP & browser extension access
  • ✓Community support & docs
Start Free
Most Popular

Pro

$79/year

Billed per year ($79/yr)

  • ✓Unlimited spaces & memories
  • ✓Unlimited devices & 20 API keys
  • ✓Cross-client memory continuity (MCP, extension, REST)
  • ✓Full per-account audit logs
  • ✓Priority support path
  • ✓Early access to new features

Business

$1950/year

Billed annually ($390/user/yr)

  • ✓Everything in Pro
  • ✓Minimum 5 users
  • ✓50 API keys for team agents
  • ✓Organization RBAC & PostgreSQL RLS isolation
  • ✓Human-in-the-loop (HITL) approval gates
  • ✓Veritas provenance and contradiction review where enabled
  • ✓Shared team spaces & role management

Enterprise

Starting at $1,500/month

Billed per year · Custom deployment

  • ✓Everything in Business
  • ✓Starting at $1,500/month (Billed per year)
  • ✓Self-host deployment (Docker + PostgreSQL + pgvector)
  • ✓Private / Dedicated Enterprise infrastructure option
  • ✓Audit review, retention, and procurement support
  • ✓Custom onboarding & dedicated support SLA options
  • ✓Security & procurement assistance
  • ✓SSO (SAML/OIDC) — Available by agreement
Talk to Sales →
ENTERPRISE DEDICATED / PRIVATE DEPLOYMENT

Need isolated cloud or air-gapped VPC infrastructure?

We offer single-tenant dedicated deployments, customer-managed cloud VPC provisioning, and custom security compliance agreements tailored for defense, health, and financial services organizations.

Request Private Quote →

Memory you can govern, not just store

Unimatrix brings enterprise-grade access control, auditability, and verification to AI context.

Row-Level Security (RLS)

PostgreSQL RLS policies enforce tenant and team isolation at the database layer, ensuring multi-tenant agents never cross memory boundaries.

Tamper-Evident Audit Logs

Memory activity and administrative access are captured in audit logs so teams can review how durable context is created, read, updated, and retired.

Human-in-the-Loop Approval

Configure approval gates for sensitive memory modifications or agent actions before they write to organizational persistent memory.

Pricing & Governance FAQ

How does Unimatrix differ from a vector database or chat history?

Vector databases only store raw embeddings and require you to build retrieval pipelines. Single-vendor chat memory is trapped inside one company's silo. Unimatrix combines persistent storage, explicit AI tool calls, a Librarian pipeline, provenance-aware memory updates, and hybrid semantic plus full-text retrieval across compatible AI tools.

What AI systems and tools can connect to Unimatrix?

Unimatrix provides native Model Context Protocol (MCP) support for Claude Desktop, Cursor, Windsurf, Continue, and VS Code. A Manifest V3 browser extension connects web LLMs including ChatGPT, Claude.ai, Gemini, Copilot, Perplexity, and Grok. Custom AI agents and backends connect via REST API or tool calls, and mobile devices connect through our native React Native app.

Is encryption client-side or server-side?

Dashboard-created memories can be encrypted in your browser before transmission using your key derivation password. Memories captured through AI integrations, MCP, browser extensions, and APIs are processed server-side so Unimatrix can organize, classify, index, and retrieve them, and then encrypted with AES-256-GCM for storage at rest.

How does Enterprise pricing work?

Enterprise deployment starts at $1,500/month on an annual contract. It includes organizational AI memory infrastructure, custom onboarding, priority support SLA options, deployment flexibility, and procurement support. Dedicated VPC/private hosting and custom compliance agreements are tailored individually by our sales team.

Can our organization self-host Unimatrix?

Yes. Self-hosting is a core Enterprise capability. You can deploy the complete stack on your own infrastructure using Docker Compose and PostgreSQL 15+ with pgvector, ensuring complete data residency and network boundary control.

How does contradiction detection work?

When new information conflicts with an existing belief, Unimatrix can preserve supersession history and provenance instead of silently overwriting past facts. Advanced Veritas contradiction analysis is surfaced where the relevant tools are enabled.