Skip to main content
POST
/
v1
/
memories
/
extract
Extract Facts
curl --request POST \
  --url https://api.memoclaw.com/v1/memories/extract \
  --header 'Content-Type: application/json' \
  --data '
{
  "messages": [
    {}
  ],
  "namespace": "<string>",
  "session_id": "<string>",
  "agent_id": "<string>"
}
'
{
  "memory_ids": [
    "<string>"
  ],
  "facts_extracted": 123,
  "facts_stored": 123,
  "facts_deduplicated": 123,
  "tokens_used": 123
}
This endpoint requires an x402 payment header.
Price: $0.01 USDC (includes LLM processing) Extract automatically identifies and stores important facts from a conversation. The LLM parses the messages and creates individual memories for each distinct fact, with automatic deduplication.

​
Request Body

​
messages
array
required
Array of conversation messages. Each message must have role (string) and content (string). Max 100 messages, 32,768 characters per message.
​
namespace
string
Namespace for extracted memories. Default: "default".
​
session_id
string
Associate extracted memories with a session.
​
agent_id
string
Associate extracted memories with an agent.

​
Response (201 Created)

​
memory_ids
string[]
UUIDs of the stored memories (includes both new and deduplicated).
​
facts_extracted
number
Total facts identified by the LLM.
​
facts_stored
number
New facts stored (not duplicates).
​
facts_deduplicated
number
Facts that matched existing memories (skipped).
​
tokens_used
number
Total tokens consumed (LLM + embeddings).

​
Example

curl -X POST https://api.memoclaw.com/v1/memories/extract \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {"role": "user", "content": "I prefer dark mode in all my apps. Also, my timezone is PST."},
      {"role": "assistant", "content": "Got it! I will remember your preferences."}
    ]
  }'
Response
{
  "memory_ids": [
    "550e8400-e29b-41d4-a716-446655440010",
    "550e8400-e29b-41d4-a716-446655440011"
  ],
  "facts_extracted": 2,
  "facts_stored": 2,
  "facts_deduplicated": 0,
  "tokens_used": 185
}
The LLM automatically assigns importance and memory type based on the content. Corrections and preferences get higher importance; observations get lower. Each fact is deduplicated against existing memories.