Introduction

Recall what matters.

Engramme is the API for Human Memory. It recalls relevant material from a user’s trained Large Memory Model while they read, write, or speak, so an application can surface memories without asking the user to search for them first. Send source material to Engramme, and the LMM training pipeline turns it into structured memories that can be recalled later with natural language.

What are Large Memory Models?

LMMs are a new class of AI models beyond LLMs. They are designed for human memory: storing, linking, and recalling experiences from a personal archive. Instead of only processing a prompt, an LMM can recall relevant memories from the material it has been trained on.

How It Works

Memorize

Send documents to the /v1/memorize endpoint. Emails, meeting notes, journal entries, and other source material can be submitted for LMM training.

LMM Trains

The training pipeline turns source material into structured memories, identifying narratives, participants, time, place, and semantic connections.

Recall

Retrieve relevant memories with /v1/recall.

Using a coding agent? We know. You’re probably not going to read every page before opening your editor.Give it /llms.txt for the short version, or /llms-full.txt when you want the full API context in one shot.

Quick Example

# Submit a document for LMM training
curl -X POST https://memorymachines-gateway-prod-btf57kda.uc.gateway.dev/v1/memorize \
  -H "x-api-key: $ENGRAMME_API_KEY" \
  -F "file=@meeting-notes.txt" \
  -F "user_name=john_doe" \
  -F "source_type=text"

Get Started

Console Setup

API Reference

SDKs

Recall what matters.

What are Large Memory Models?

How It Works

Quick Example

Next Steps

Quick Start

API Reference

​Recall what matters.

​What are Large Memory Models?

​How It Works

​Quick Example

​Next Steps

Quick Start

API Reference

Recall what matters.

What are Large Memory Models?

How It Works

Quick Example

Next Steps