Your agent forgets everything. Mengram fixes that.

Your agent learns from every session. Facts, workflows, and procedures — auto-saved, auto-recalled, auto-improved. Works natively in 23 languages.

$ pip install mengram-ai

✓ Successfully installed mengram-ai

from mengram import Mengram
m = Mengram("om-..

No credit card. No trial. Free forever tier with 40 memories/mo.

How it works

1

You chat with any AI

Use ChatGPT, Claude Desktop, Cursor, Perplexity — any AI you prefer. Mengram connects via MCP or API.

2

Mengram extracts 3 memory types

Semantic — facts, preferences, skills. Episodic — events, discussions, decisions. Procedural — workflows, processes, habits.

3

Every AI knows you deeply

One API call returns a Cognitive Profile — a ready-to-use system prompt from all 3 memory types. Zero effort personalization.

Before Mengram / After Mengram

Replace your entire RAG pipeline with 3 lines of code.

✕ Traditional RAG Pipeline

from langchain.embeddings import OpenAIEmbeddings
from langchain.vectorstores import Pinecone
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.chains import RetrievalQA
import pinecone

pinecone.init(api_key="...", environment="...")
embeddings = OpenAIEmbeddings()
splitter = RecursiveCharacterTextSplitter(chunk_size=500)
chunks = splitter.split_documents(docs)
vectorestore = Pinecone.from_documents(chunks, embeddings)
retriever = vectorstore.as_retriever(search_kwargs={"k": 5})
chain = RetrievalQA.from_chain_type(llm=llm, retriever=retriever)
result = chain.run("What does Ali prefer?")

15 lines · 3 API keys · manual chunking

✓ With Mengram

from mengram import Mengram
m = Mengram(api_key="om-...")
results = m.search("What does Ali prefer?")

3 lines · 1 API key · zero config

Simple, predictable pricing

Start free. Upgrade when you need more.

Free

$0

Try it out — no credit card needed.

• 40 memory adds / month
• 200 searches / month
• 3 agent runs
• 3 sub-users
• 20 req/min rate limit
• Vector search (no reranking)
• No procedure evolution
• No smart triggers

Starter

$5 / month

For personal projects and indie developers.

• 100 memory adds / month
• 500 searches / month
• 10 agent runs
• 10 sub-users
• 60 req/min rate limit
• Vector search (no reranking)
• 2 webhooks
• 1 team

Pro

$19 / month

For production apps and power users.

• 1,000 memory adds / month
• 10,000 searches / month
• 50 agent runs
• 50 sub-users
• 120 req/min rate limit
• LLM-powered reranking
• Procedure evolution
• Smart triggers
• 10 webhooks
• 5 teams

Growth

$59 / month

For scaling apps that need more volume.

• 3,000 memory adds / month
• 20,000 searches / month
• Unlimited agent runs
• 100 sub-users
• 200 req/min rate limit
• Cohere cross-encoder reranking
• Procedure evolution
• Smart triggers
• 25 webhooks
• 10 teams

Business

$99 / month

For teams and high-volume applications.

• 8,000 memory adds / month
• 30,000 searches / month
• Unlimited agent runs
• Unlimited sub-users
• 300 req/min rate limit
• Cohere cross-encoder reranking
• Procedure evolution
• Smart triggers
• 50 webhooks
• Unlimited teams

Enterprise

Custom

For organizations with custom requirements.

• Custom memory & search limits
• Dedicated infrastructure
• Custom rate limits
• SSO & access controls
• Priority support & SLA
• Custom integrations
• Data residency options
• On-premise deployment

What makes Mengram different

Others store facts. Mengram remembers like a human brain.

Memory Agents

Curator cleans contradictions. Connector finds hidden patterns. Digest gives weekly briefs. Runs autonomously.

Multi-User Isolation

One API key, many users. Pass user_id to scope memories per end-user. Each user gets their own isolated facts, events, workflows, and cognitive profile.

Mengram vs Mem0 vs Supermemory

Others store facts. Mengram remembers experiences and learns workflows.

Get started in 60 seconds

Connect Mengram to your AI tools via MCP, Python, or JavaScript SDK.

1

Install mengram

Copypip install mengram-ai

2

Use in your app

from mengram import Mengram

m = Mengram(api_key="om-...")

# Save — auto-extracts facts, events, workflows
m.add([
    {"role": "user", "content": "Fixed OOM with Redis cache"},
    {"role": "assistant", "content": "Got it."},
])

# Unified search — all 3 memory types
results = m.search_all("database issues")
# → {semantic: [...], episodic: [...], procedural: [...]}

# Cognitive Profile — instant personalization
profile = m.get_profile()
# → ready system prompt for any LLM

# Multi-user isolation — one API key, many users
m.add([...], user_id="alice")
m.search_all("prefs", user_id="alice")  # only Alice's data