Add Memory

curl --request POST \ --url 'https://{workspaceId}.cn-beijing.maas.aliyuncs.com/api/v2/apps/memory/add' \ --header 'Authorization: Bearer <YOUR_API_KEY>' \ --header 'Content-Type: application/json' \ --data ' { "user_id": "<string>", "messages": [ { "role": "user", "content": "<string>" } ], "custom_content": "<string>", "profile_schema": "<string>", "memory_library_id": "<string>", "meta_data": {} } '

Stores user conversations as factual memories, automatically extracting key information. Also supports directly specifying custom content for storage. To extract a user profile at the same time, pass profile_schema.

Before calling, ensure you have obtained an API Key. See Authentication.

For long conversations where extraction takes time, use Add Memory Async, which runs extraction in the background and reports status and results by event ID.

Request parameters

Parameter	Type	Required	Description
`user_id`	string	Yes	Memory entity ID, used to identify the owning object, max 64 characters
`messages`	array	Yes*	Conversation message list, up to 50 entries. Each contains `role` (user/assistant) and `content`
`custom_content`	string	Yes*	Custom content, max 512 characters. Mutually exclusive with `messages`
`profile_schema`	string	No	Profile schema ID. If omitted, no user profile is extracted — only factual memories are written
`memory_library_id`	string	No	Memory library ID. If not provided, the default library is used
`project_id`	string	No	Factual memory rule ID. If not provided, the default rule is used
`meta_data`	object	No	User-defined information

messages and custom_content are mutually exclusive. When custom_content is provided, messages is ignored.

Call succeeded but profile attributes are empty? Factual memories and user profiles are extracted by two independent pipelines. Without profile_schema, the API returns only memory_nodes and profile attributes stay empty. Call again with the profile schema ID — get it from ListProfileSchemas, or from Memory rules > Profile rules on the memory detail page in the console.

Response

Field	Type	Description
`request_id`	string	Request ID
`memory_nodes`	array	List of changed factual memories
`memory_nodes[].memory_node_id`	string	Factual memory ID
`memory_nodes[].content`	string	Extracted memory content
`memory_nodes[].event`	string	Operation type: ADD / UPDATE / DELETE
`memory_nodes[].old_content`	string	Content before update, only valid when event is UPDATE

Code examples

Replace {workspaceId} with your actual workspace ID.

curl -X POST https://{workspaceId}.cn-beijing.maas.aliyuncs.com/api/v2/apps/memory/add \
  --header "Authorization: Bearer $DASHSCOPE_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
    "messages": [
      {"role": "user", "content": "每天上午9点提醒我喝水"},
      {"role": "assistant", "content": "好的，已记录"},
      {"role": "user", "content": "明天10点提醒我整理会议纪要。"}
    ],
    "user_id": "user_001"
  }'

# response
{
  "memory_nodes": [
    {"content": "用户每天上午9点需要喝水提醒", "event": "ADD", "memory_node_id": "50b46e97..."},
    {"content": "用户明天10点需要整理会议纪要", "event": "ADD", "memory_node_id": "381cc694..."}
  ],
  "request_id": "bec69131-627c-..."
}

After writing, use SearchMemory to search memories.

Add Memory

Request parameters

Response

Code examples

Authorizations

Body

Response

​Request parameters

​Response

​Code examples

Authorizations

Body

Response

Request parameters

Response

Code examples