Creates a new collection with the specified name and schema. Collections are containers for documents that share a common structure.
Method Signature
collections().create(
name: string,
schema: Record<string, FieldSpec>
): Promise<Collection>
collections().create(
collection_name: str,
schema: Mapping[str, FieldSpec]
) -> Collection
Parameters
The name of the collection to create. Must be unique within the project.
schema
Record<string, FieldSpec>
required
The schema definition for the collection. Maps field names to their specifications, including data types and indexes.
Returns
Returns a Collection object containing metadata about the newly created collection.Organization ID that owns the collection
Project ID that contains the collection
schema
Record<string, CollectionFieldSpec>
Schema definition for the collection fields
Region where the collection is stored
Examples
Basic Text Collection
import { Client } from "topk-js";
import { text, int } from "topk-js/schema";
const client = new Client({
apiKey: process.env.TOPK_API_KEY,
region: "us-east-1"
});
const collection = await client.collections().create("books", {
title: text().required(),
author: text(),
published_year: int()
});
console.log(`Created collection: ${collection.name}`);
from topk_sdk import Client
from topk_sdk.schema import text, int
client = Client(
api_key=os.environ["TOPK_API_KEY"],
region="us-east-1"
)
collection = client.collections().create("books", schema={
"title": text().required(),
"author": text(),
"published_year": int()
})
print(f"Created collection: {collection.name}")
Vector Search Collection
import { Client } from "topk-js";
import { text, f32Vector, vectorIndex } from "topk-js/schema";
const client = new Client({
apiKey: process.env.TOPK_API_KEY,
region: "us-east-1"
});
const collection = await client.collections().create("products", {
name: text().required(),
description: text(),
embedding: f32Vector({ dimension: 1536 }).index(
vectorIndex({ metric: "cosine" })
)
});
console.log(`Created collection with vector index: ${collection.name}`);
from topk_sdk import Client
from topk_sdk.schema import text, f32_vector, vector_index
client = Client(
api_key=os.environ["TOPK_API_KEY"],
region="us-east-1"
)
collection = client.collections().create("products", schema={
"name": text().required(),
"description": text(),
"embedding": f32_vector(dimension=1536).index(
vector_index(metric="cosine")
)
})
print(f"Created collection with vector index: {collection.name}")
Keyword Search Collection
import { Client } from "topk-js";
import { text, int, keywordIndex } from "topk-js/schema";
const client = new Client({
apiKey: process.env.TOPK_API_KEY,
region: "us-east-1"
});
const collection = await client.collections().create("articles", {
title: text().required().index(keywordIndex()),
content: text().index(keywordIndex()),
views: int()
});
console.log(`Created collection with keyword indexes: ${collection.name}`);
from topk_sdk import Client
from topk_sdk.schema import text, int, keyword_index
client = Client(
api_key=os.environ["TOPK_API_KEY"],
region="us-east-1"
)
collection = client.collections().create("articles", schema={
"title": text().required().index(keyword_index()),
"content": text().index(keyword_index()),
"views": int()
})
print(f"Created collection with keyword indexes: {collection.name}")
Semantic Search Collection
import { Client } from "topk-js";
import { text, semanticIndex } from "topk-js/schema";
const client = new Client({
apiKey: process.env.TOPK_API_KEY,
region: "us-east-1"
});
const collection = await client.collections().create("documents", {
title: text().required(),
content: text().index(
semanticIndex({ model: "cohere/embed-v4" })
)
});
console.log(`Created collection with semantic index: ${collection.name}`);
from topk_sdk import Client
from topk_sdk.schema import text, semantic_index
client = Client(
api_key=os.environ["TOPK_API_KEY"],
region="us-east-1"
)
collection = client.collections().create("documents", schema={
"title": text().required(),
"content": text().index(
semantic_index(model="cohere/embed-v4")
)
})
print(f"Created collection with semantic index: {collection.name}")
Once created, a collection’s schema cannot be modified. Plan your schema carefully before creating the collection.
