OthersMemory

Mem0

Mem0 is a self-improving memory layer for LLM applications, enabling personalized AI experiences. Use it with Respan to add memory to your AI product and route LLM calls through the Respan gateway for full observability over every memory add, search, and generation.

Create an account at platform.respan.ai and grab an API key. For gateway, also add credits or a provider key.

Run npx @respan/cli setup to set up with your coding agent.

Mem0 doesn’t expose its own SDK-level tracing instrumentor; route all calls through the Respan gateway to get full observability automatically.

Setup

1

Prerequisites

2

Install Mem0

pip install mem0ai
3

Set environment variables

export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"
export MEM0_API_KEY="YOUR_MEM0_API_KEY"
4

Add memory with Mem0 SDK

Configure Mem0’s LLM provider to use the Respan gateway.

import os
from mem0 import Memory
config = {
    "llm": {
        "provider": "openai",
        "config": {
            "model": "gpt-4.1-nano",
            "temperature": 0.0,
            "api_key": os.environ["RESPAN_API_KEY"],
            "openai_base_url": "https://api.respan.ai/api/",
        },
    }
}
m = Memory.from_config(config_dict=config)
result = m.add(
    "I like to take long walks on weekends.",
    user_id="alice",
    metadata={"category": "hobbies"},
)
print(result)

Add memory with the OpenAI SDK

Use the OpenAI SDK pointed at the Respan gateway and pass mem0_params via extra_body to add or search memories alongside the LLM call.

import os
from openai import OpenAI
client = OpenAI(
    api_key=os.environ["RESPAN_API_KEY"],
    base_url="https://api.respan.ai/api/",
)
messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Context: I like eating carrots, and I like to play basketball."},
]
response = client.chat.completions.create(
    model="gpt-4.1-nano",
    messages=messages,
    extra_body={
        "mem0_params": {
            "user_id": "user_1",
            "org_id": "org_1",
            "api_key": os.environ["MEM0_API_KEY"],
            "add_memories": {"messages": messages},
        }
    },
)
print(response.choices[0].message.content)

Search memories

Pass search_memories to retrieve relevant memories before generation.

response = client.chat.completions.create(
    model="gpt-4.1-nano",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Based on {{mem0_search_memories_response}}, what is the user's favorite food?"},
    ],
    extra_body={
        "mem0_params": {
            "user_id": "user_1",
            "org_id": "org_1",
            "api_key": os.environ["MEM0_API_KEY"],
            "search_memories": {
                "query": "What is the user's favorite food?",
                "top_k": 1,
                "fields": ["text"],
                "rerank": True,
                "keyword_search": True,
                "filter_memories": True,
            },
        }
    },
)

Use the Respan SDK for type-safe params

pip install respan-sdk
import os
from openai import OpenAI
from respan_sdk.respan_types.services_types.mem0_types import (
    Mem0Params, AddMemoriesParams, SearchMemoriesParams,
)
client = OpenAI(
    api_key=os.environ["RESPAN_API_KEY"],
    base_url="https://api.respan.ai/api/",
)
memory_msgs = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Context: I like eating carrots."},
]
response = client.chat.completions.create(
    model="gpt-4.1-nano",
    messages=memory_msgs,
    extra_body={
        "mem0_params": Mem0Params(
            user_id="user_1",
            org_id="org_1",
            api_key=os.environ["MEM0_API_KEY"],
            add_memories=AddMemoriesParams(messages=memory_msgs),
        ).model_dump(exclude_none=True),
    },
)