Launching DolphinBench - a memory benchmark for how AI agents act on real tasks.

Launching DolphinBench - a memory benchmark for how AI agents act on real tasks.

Launching DolphinBench - a memory benchmark for how AI agents act on real tasks.

Backed by

Y Combinator Logo

Combinator

AI memory that persists across sessions and agents

AI memory that persists across sessions and agents

Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.

Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.

Setup for Agent

Home

home_secondary_agentsetup

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
# Step 1 — Install the SDK. Run this in your terminal (not in Python):## pip install mem0ai## Step 2 — Save this file as mem0_quickstart.py and run: python mem0_quickstart.pyimport osfrom mem0 import MemoryClient # Set your API key (get one at https://app.mem0.ai)client = MemoryClient(api_key=os.getenv("MEM0_API_KEY", "your-api-key-here")) # Use the same user ID when adding and searchingUSER_ID = "user123" # Add a memorymessages = [ {"role": "user", "content": "I'm a vegetarian and allergic to nuts."}, {"role": "assistant", "content": "Got it! I'll remember your dietary preferences."},]client.add(messages, user_id=USER_ID) # Search memoriesresults = client.search( "What are my dietary restrictions?", filters={"user_id": USER_ID})print(results)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
# Step 1 — Install the SDK. Run this in your terminal (not in Python):## pip install mem0ai## Step 2 — Save this file as mem0_quickstart.py and run: python mem0_quickstart.pyimport osfrom mem0 import MemoryClient # Set your API key (get one at https://app.mem0.ai)client = MemoryClient(api_key=os.getenv("MEM0_API_KEY", "your-api-key-here")) # Use the same user ID when adding and searchingUSER_ID = "user123" # Add a memorymessages = [ {"role": "user", "content": "I'm a vegetarian and allergic to nuts."}, {"role": "assistant", "content": "Got it! I'll remember your dietary preferences."},]client.add(messages, user_id=USER_ID) # Search memoriesresults = client.search( "What are my dietary restrictions?", filters={"user_id": USER_ID})print(results)

150,000+

150,000+

150,000+

150,000+

150,000+

150,000+

Developers build with Mem0

Built for <developers> who want proof,
not promises

Built for <developers> who want proof,
not promises

Built for <developers> who want proof,
not promises

Mem0 gives agents persistent memory without pipeline changes. Less redundant context, lower token costs, measurably faster responses.

Mem0 gives agents persistent memory without pipeline changes. Less redundant context, lower token costs, measurably faster responses.

Memory Compression Engine

Automatically condenses chat history into compact memories that cut tokens and latency while keeping the right context.

Memory Compression Engine

Automatically condenses chat history into compact memories that cut tokens and latency while keeping the right context.

How it works

Add anything. Mem0 learns

Add

Input data in seconds with
no config or boilerplate

Learn

Mem0 extracts and

updates memories

Retrieve

Mem0 retrieves key memories as users interact

Add

Input data in seconds with no config or boilerplate

Learn

Mem0 extracts and

updates memories

Retrieve

Mem0 retrieves key memories as users interact

AI memory that adapts
to your domain

Mem0 helps AI remember what matters.

Mem0 helps AI remember what matters.

Smart Patient Care Assistant

Remembers patient history, allergies, and treatment preferences across visits therefore providing personalized care that improves with every interaction.

Chronic Condition Companion

Learns what works (and what doesn’t) for the patient over time, offering thoughtful reminders and insights tailored to each patient’s journey.

Therapy Progress Tracker

Builds on previous sessions to deliver consistent, context-aware mental health support. Creates trust through conversations that remember what matters to each patient.

HEALTHCARE

Smart Patient Care Assistant

Remembers patient history, allergies, and treatment preferences across visits therefore providing personalized care that improves with every interaction.

Education

Adaptive Learning Tutor

Adapts to each student's pace and learning style, remembering what works best. Transforms one-size-fits-all education into personalized learning that evolves with every lesson.

Sales & CRM

Sales Assistant with Persistent Context

Track every interaction, objection, and milestone across long sales cycles ensuring reps have instant recall at every touchpoint.

New Algorithm

Benchmarking Mem0

Single-pass hierarchical distillation. Multi-signal retrieval. Benchmarked across LoCoMo, LongMemEval, and BEAM.

Single-pass hierarchical distillation. Multi-signal retrieval. Benchmarked across LoCoMo, LongMemEval, and BEAM.

mem0 benchmark scores

Built for enterprise
Designed for control

Memory at scale is infrastructure. Mem0 gives enterprise teams governance, reliability, and full observability so engineers spend time building, not recovering lost context.

Memory at scale is infrastructure. Mem0 gives enterprise teams governance, reliability, and full observability so engineers spend time building, not recovering lost context.

Governance

SOC 2, HIPAA, BYOK, zero-trust. Your data stays yours.

Portable

Kubernetes, private cloud, or air-gapped. Same API everywhere.

Auditable

Every read and write logged. Know what, who, and when.

We take security and privacy seriously. Mem0 is SOC 2 (Type 1), HIPAA, and GDPR compliant, ensuring your data is protected with industry-standard safeguards at every step.

We take security and privacy seriously. Mem0 is SOC 2 (Type 1), HIPAA, and GDPR compliant, ensuring your data is protected with industry-standard safeguards at every step.

AICPA SOC
HIPA Compliant
GDPR logo