Vectorize

What is Vectorize?

Persistent, model-agnostic memory for agents.

Who is Vectorize best for?

Vectorize is best for developers building AI agents that must remember each user across sessions and learn from mistakes.

• Agents with returning users

  Each user gets separate memory of preferences, history and decisions that carries over between sessions.

• Teams that want to self-host

  Hindsight is MIT licensed and runs from one Docker command with its own PostgreSQL, with no usage limits.

• Coding agents and MCP clients

  A built-in MCP server gives Claude Code, Cursor and other MCP clients remember, recall and reflect tools.

• Apps that switch models

  The memory layer works with any LLM, so changing models does not wipe what the agent has learned.

How much does Vectorize cost?

Vectorize's Hindsight agent memory is free to self-host under the MIT license, while the managed Hindsight Cloud has no monthly fee and bills each operation, such as $10 per million tokens to retain memories and $0.75 per million tokens to recall them.

Enterprise is priced on request and adds bring-your-own-cloud or on-premises deployment, SSO and role-based access, and support up to 24x7 with a 30-minute response time. Other Cloud rates: Iris document extraction $7.50 per million tokens, mental model retrieval $0.25 per million tokens and refresh $0.05 per call. Cloud credits come in packs of $10, $25, $50 or $100, or any amount from $5 to $1,000, at $1 per credit, and volume discounts are available on request. The self-hosted version has no usage limits. Full pricing on vectorize.io. We checked these prices on 28 Sep 2026.

How do readers rate Vectorize?

No one has rated Vectorize yet. Be the first.

Is Vectorize open source?

Yes. Vectorize's code is public under the MIT licence, so you can run it on your own server.

Host it on a serverRailway and Elestio set it up on a server for you with a one-click install.

Hosting with

What are the best Vectorize alternatives?

Vectorize is an open source alternative to Zep. Honcho, MemPalace and Letta do the same job and are open source too.

ZepFree

Temporal memory and context graphs for agents.

HonchoOpen Source

Memory library for agents that remember users long term.

MemPalaceOpen Source

Open-source local memory for AI agents such as Claude Code, saving chats verbatim for search.

LettaOpen Source

Persistent agents with memory that carries across models.

Questions about Vectorize