Skip to content
OctaMem
Start free

Memory at the scale your organization runs.

Three plans for teams, businesses, and regulated enterprises. Storage scales linearly. You bring your own LLM keys.

Plan fit

Price the shape of your memory layer.

Choose the amount of structured memory and the number of seats. The fit guide points you toward the tier that usually matches.

250 GB
25

Likely fit

Team

Starting point

$1,200 / month

Why

Best for the first production agent team.

  • Team

    Team.

    For product teams shipping agents

    $1,200/ month

    250 GB structured memory

    Up to 25 seats

    • Semantic, episodic, and procedural memory types
    • REST API, MCP server, Python and JavaScript SDKs
    • Audit log of every read and write
    • Role-based scopes (department, project, agent)
    • Email + chat support, 24h response
    Book a discovery call

    Trial includes 50 GB. Card required after 14 days.

  • Our recommendation

    Business

    Business.

    For organizations standardizing on OctaMem

    $4,800/ month

    1 TB structured memory

    Up to 100 seats

    • Everything in Team
    • SSO / SAML, SCIM provisioning
    • Sandbox environments per project
    • Custom retention policies and BYO-KMS
    • Slack Connect support, 4h response
    • Quarterly security review with our team
    Book a discovery call

    Annual billing available; ask about prepay discounts.

  • Enterprise

    Enterprise.

    For regulated, high-stakes deployments

    Custom

    Multi-TB structured memory

    Unlimited seats

    • Everything in Business
    • Single-tenant cloud, VPC, or on-premise deployment
    • Customer-managed encryption keys (BYO-KMS)
    • Dedicated solutions engineer and security review
    • Vertical clouds for Healthcare, Finance, Defense, and Legal
    • DPA, BAA, and custom contracts; SOC 2 and ISO on roadmap
    • Priority incident response with negotiated SLA
    Book a discovery call

    DPA, BAA, and custom contracts available.

§ 02Frequently asked

  • What counts as memory storage?

    Structured memory records (semantic, episodic, procedural) plus their indices and audit lineage. Source files like PDFs, decks, and spreadsheets you ingest get parsed into typed memory and aren't double-counted.

  • What about AI usage costs?

    You bring your own LLM keys (OpenAI, self-hosted). We run the memory; you manage the model budget.

  • Is on-premise deployment really available?

    Yes. Available on Enterprise. We ship a self-contained deployment with your team's hardware specs. Single-tenant cloud and customer VPC are also options.

  • Do you sign DPAs and BAAs?

    DPA at Business and above. BAA on Enterprise for healthcare deployments. Both available for review during procurement.

Still have questions?

Talk to sales