Create Index - CyborgDB Docs

Creates and returns an EncryptedIndex instance based on the provided configuration.

async createIndex(
    indexName: string, 
    indexKey: Uint8Array, 
    indexConfig: IndexIVFPQModel | IndexIVFFlatModel | IndexIVFModel,
    embeddingModel?: string
): Promise<EncryptedIndex>

Parameters

Parameter	Type	Description
`indexName`	`string`	Name of the index to create. Must be unique within the CyborgDB service.
`indexKey`	`Uint8Array`	32-byte encryption key for the index, used to secure the index data.
`indexConfig`	`IndexIVFPQModel` \| `IndexIVFFlatModel` \| `IndexIVFModel`	Configuration object specifying the index type (`ivf`, `ivfpq`, or `ivfflat`) and relevant parameters such as `dimension`, `nLists`, `pqDim`, and `pqBits`.
`embeddingModel`	`string`	(Optional) Name of the embedding model used to auto-generate embeddings on the server. Defaults to `undefined`, no auto-generation.

For more info on auto-generating embeddings, refer to Auto-Generate Embeddings.

Returns

Promise<EncryptedIndex>: A Promise that resolves to an instance of the newly created encrypted index.

Exceptions

Error

Throws if the index name already exists on the server.
Throws if the index configuration is invalid or missing required parameters.
Throws if the encryption key is not exactly 32 bytes.
Throws if the embedding model is not supported by the server.

Network/API Errors

Throws if the API request fails due to network issues.
Throws if the server returns an HTTP error status.
Throws if authentication fails (invalid API key).

Example Usage

Basic IVFFlat Index Creation

import { Client, IndexIVFFlatModel } from 'cyborgdb';

const client = new Client('http://localhost:8000', 'your-api-key');

const indexName = "my_vector_index";
const indexKey = crypto.getRandomValues(new Uint8Array(32)); // Generate secure 32-byte key

const indexConfig: IndexIVFFlatModel = {
    dimension: 768,
    nLists: 1024,
    metric: 'cosine'
};

try {
    const index = await client.createIndex(
        indexName, 
        indexKey, 
        indexConfig
    );
    console.log('Index created successfully:', index);
} catch (error) {
    console.error('Failed to create index:', error.message);
}

IVFPQ Index with Embedding Model

import { Client, IndexIVFPQModel } from 'cyborgdb';

const client = new Client('http://localhost:8000', 'your-api-key');

const indexName = "semantic_search_index";
const indexKey = crypto.getRandomValues(new Uint8Array(32));

const indexConfig: IndexIVFPQModel = {
    dimension: 1536,
    nLists: 2048,
    metric: 'cosine',
    pqDim: 64,
    pqBits: 8
};

try {
    const index = await client.createIndex(
        indexName,
        indexKey,
        indexConfig,
        'all-MiniLM-L6-v2'  // OpenAI embedding model
    );
    
    console.log('IVFPQ index with embeddings created successfully');
} catch (error) {
    console.error('Index creation failed:', error.message);
}

⌘I