Knowledge Memory Pricing Docs
/v1/memory/write
Product 02 / 02
Per-user isolated
Semantic recall
Memory Product · Free tier live

Give your AI a memory.

Persistent semantic memory for AI agents. Write once, retrieve forever.
Organize by project. Recall by meaning, not by keyword.

Live demo

Watch your AI remember.

Your AI stores context from past conversations. Later, it recalls exactly what it needs - project decisions, user preferences, anything you told it before.

Coding Agent
Ask a question...
How it works

Three verbs. One HTTP endpoint.

No vector DB to run. No embedding to tune. Just /write, /search, and (when needed) /update. Your agent handles the reasoning. We handle the recall.

1 WRITE

Capture the context.

POST any content to /v1/memory/write. Returns immediately with a memory id while our engine processes it in the background.

2 INDEX

Processed in the background.

Indexing happens silently in the background. No wait, no blocking - just poll the status endpoint until it's ready.

3 RECALL

Search by meaning.

Your agent asks a question, scoped to a project or across all of them. We return the most relevant memories, ordered by semantic similarity.

Project organization

One agent. Many projects.

A coding agent works across multiple repositories. Memory stays scoped to the project the agent is working on, with a single label passed on every call. Cross-project search when you need it.

01 / Scope

Organize by project.

Every memory lives under a project name. No more mixing contexts between repositories, clients, or sessions.

02 / Pass-through

AI passes the project name.

Your agent sends the project label on every write and search. Perfect for coding assistants that switch between repos.

03 / Cross-project

Global recall when needed.

Omit the project on search and Memory looks across everything the user has ever stored - useful for personal preferences that span projects.

04 / Lifecycle

Bulk rename & delete.

Renamed your repo? Rename the project in one call. Archived a client? Delete the whole project and its memories together.

Why Ragionex Memory

Persistent memory, without the infrastructure.

Agents need memory that survives across sessions, users, and restarts. Ragionex Memory is the storage and retrieval layer - your agent's model does the reasoning on top.

01 / Isolation

Per-user isolation.

One agent, many users - each user's memories stay fully scoped and invisible to the rest. No cross-contamination, no leakage.

02 / Async

Background indexing.

Write returns in milliseconds. The engine indexes it the right way while your agent moves on to the next turn.

03 / Recall

Semantic, not literal.

Ask "what does the user prefer?" and get back facts stored as "I hate dark chocolate." Semantic search finds the right memory even when the query wording is completely different from how it was stored.

04 / MCP

MCP compatible.

Expose Memory to your agent as an MCP tool. Claude, Cursor, Windsurf, and any MCP-compatible client can read and write memories with one config line.

Get started free

Grab your API key.

Memory Free includes 1,000 memories, 100 projects, and cross-project search - everything you need to start, free forever.
No credit card. Data stored in EU.
Power user? See Pro pricing →