Memory: Create v1

The create memory API endpoint allows you to create a new memory on Langbase dynamically with API. This endpoint requires a User or Org API key.

Generate a User/Org API key

You will need to generate an API key to authenticate your requests. For more information, visit the User/Org API key documentation.

POST/v1/memory

Create a new memory

Create a new memory by sending the memory data inside the request body.

Headers

Name
Content-Type
Type
string
Required
Required
Description
Request content type. Needs to be application/json.
Name
Authorization
Type
string
Required
Required
Description
Replace <YOUR_API_KEY> with your user/org API key.

Body Parameters

Name
name
Type
string
Required
Required
Description
Name of the memory.
Name
description
Type
string
Description
Short description of the memory.
Default: ''

Optional Parameters

Name
embedding_model
Type
string
Description
Name of the embedding model to use for the memory.
Default: openai:text-embedding-3-large
Supported models:
- openai:text-embedding-3-large
- cohere:embed-v4.0
- cohere:embed-multilingual-v3.0
- cohere:embed-multilingual-light-v3.0
- google:text-embedding-004
Name
chunk_size
Type
number
Description
Maximum number of characters in a single chunk.
Default: 10000
Maximum: 30000
Appropriate value
Cohere has a limit of 512 tokens (1 token ~= 4 characters in English). If you are using Cohere models, adjust the chunk_size accordingly. For most use cases, default values should work fine.
Name
chunk_overlap
Type
number
Description
Number of characters to overlap between chunks.
Default: 2048
Maximum: Less than chunk_size
Name
top_k
Type
number
Description
Number of chunks to return.
Default: 10
Minimum: 1 Maximum: 100

Usage example

Install the SDK

npm i langbase

Environment variables

.env file

LANGBASE_API_KEY="<YOUR_API_KEY>"

Create memory

Create Memory

POST

/v1/memory

curl https://api.langbase.com/v1/memory \
-H 'Content-Type: application/json' \
-H "Authorization: Bearer <YOUR_API_KEY>" \
-d '{
  "name": "knowledge-base",
  "description": "An AI memory for storing company internal docs.",
  "embedding_model": "openai:text-embedding-3-large",
  "chunk_size": 10000,
  "chunk_overlap": 2048,
  "top_k": 10
}'

Response

Name
Memory
Type
object
Description
The response object returned by the API endpoint.
Memory
interface Memory { name: string; description: string; owner_login: string; url: string; chunk_size: number; chunk_overlap: number; embedding_model: | 'openai:text-embedding-3-large' | 'cohere:embed-multilingual-v3.0' | 'cohere:embed-multilingual-light-v3.0' | 'google:text-embedding-004'; }
- Name
  name
  Type
  string
  Description
  Name of the memory.
- Name
  description
  Type
  string
  Description
  Description of the AI memory.
- Name
  owner_login
  Type
  string
  Description
  Login of the memory owner.
- Name
  url
  Type
  string
  Description
  Memory studio URL.
- Name
  embedding_model
  Type
  string
  Description
  The embedding model used by the AI memory.
  
  openai:text-embedding-3-large
  
  cohere:embed-multilingual-v3.0
  
  cohere:embed-multilingual-light-v3.0
  
  google:text-embedding-004

API Response

{
  "name": "knowledge-base",
  "description": "Advanced memory with multilingual support.",
  "chunk_size": 10000,
  "chunk_overlap": 2048,
  "owner_login": "user123",
  "url": "https://langbase.com/memorysets/user123/knowledge-base",
  "embedding_model": "openai:text-embedding-3-large",
}

Memory API