Mem0MemoryRetriever

Retrieve memories from a Mem0MemoryStore as a list of ChatMessage objects for use in AI agent pipelines.

Key Features

• Fetches relevant memories from Mem0 memory storage before passing context to a language model or agent.
• Returns memories as ChatMessage system messages, ready to be injected into a chat history.
• Supports scoping by user_id, run_id, agent_id, or app_id to serve multiple users from one pipeline.
• Pass query=None to retrieve all memories within the given scope instead of doing a relevance search.
• Supports Haystack-style filters combinable with entity ID scoping.
• Each returned message carries Mem0 retrieval metadata (memory_id, user_id, score, timestamps) under meta["mem0"].

Configuration

1. Drag the Mem0MemoryRetriever component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   • Configure the Mem0MemoryStore with your Mem0 API credentials.
   • Set top_k to control the maximum number of memories to retrieve.

Connections

Mem0MemoryRetriever receives a text query, typically from the Input component, and optional scoping parameters such as user_id. It returns a list of ChatMessage objects through its memories output. Connect the output to a component that accepts chat messages, such as a chat generator or an agent.

Source Code

To check this component's source code, open retriever.py in the Haystack Core Integrations repository.

Usage Examples

Basic Configuration

  Mem0MemoryRetriever:
    type: haystack_integrations.components.retrievers.mem0.retriever.Mem0MemoryRetriever
    init_parameters:
      top_k: 5
      memory_store:
        type: haystack_integrations.memory_stores.mem0.memory_store.Mem0MemoryStore
        init_parameters:
          api_key: YOUR_MEM0_API_KEY

Using the Component in a Pipeline

# haystack-pipeline
components:
  Mem0MemoryRetriever:
    type: haystack_integrations.components.retrievers.mem0.retriever.Mem0MemoryRetriever
    init_parameters:
      top_k: 5
      memory_store:
        type: haystack_integrations.memory_stores.mem0.memory_store.Mem0MemoryStore
        init_parameters:
          api_key: YOUR_MEM0_API_KEY

connections: []

max_runs_per_component: 100

metadata: {}

inputs:
  query:
  - Mem0MemoryRetriever.query
  user_id:
  - Mem0MemoryRetriever.user_id

outputs:
  memories: Mem0MemoryRetriever.memories

Parameters

Inputs

Parameter	Type	Description
query	Optional[str]	Text query used to search for relevant memories. Pass None to retrieve all memories within the given scope.
user_id	Optional[str]	User ID to scope the search.
run_id	Optional[str]	Run ID to scope the search.
agent_id	Optional[str]	Agent ID to scope the search.
app_id	Optional[str]	App ID to scope the search.
filters	Optional[Dict[str, Any]]	Haystack-style filters to apply. Combined with entity ID parameters when both are provided.
top_k	Optional[int]	Maximum number of memories to return. Overrides the init-time default.

Outputs

Parameter	Type	Description
memories	List[ChatMessage]	Retrieved memories as ChatMessage system messages. User-provided Mem0 metadata is included in each message's meta. Mem0 retrieval fields such as memory_id, user_id, score, and timestamps are available under meta["mem0"].

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
memory_store	Mem0MemoryStore		An instance of Mem0MemoryStore to retrieve memories from.
top_k	int	5	Default maximum number of memories to return per query.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Default	Description
query	Optional[str]		Text query to search for relevant memories. Pass None to retrieve all memories in scope.
user_id	Optional[str]	None	User ID to scope the search.
run_id	Optional[str]	None	Run ID to scope the search.
agent_id	Optional[str]	None	Agent ID to scope the search.
app_id	Optional[str]	None	App ID to scope the search.
filters	Optional[Dict[str, Any]]	None	Haystack-style filters. Combined with entity ID parameters when both are provided.
top_k	Optional[int]	None	Override the init-time top_k setting.

Related Information

• Mem0MemoryWriter
• Pipeline Components