Qdrant CreateCollection

Qdrant CreateCollection

Certified

Create a collection in Qdrant

Creates a new collection with a dense vector of the given dimension and distance metric. By default the dense vector is unnamed. Set vectorName to give it a name and sparseVectors to add named sparse vector spaces, which is the layout HybridQuery needs. Collections with several named dense vectors should be created via the Qdrant API or CLI.

yaml
type: io.kestra.plugin.qdrant.CreateCollection

Create a Qdrant collection with 1536 dimensions and Cosine distance

yaml
id: qdrant_create_collection
namespace: company.team

tasks:
  - id: create_collection
    type: io.kestra.plugin.qdrant.CreateCollection
    host: "{{ secret('QDRANT_HOST') }}"
    apiKey: "{{ secret('QDRANT_API_KEY') }}"
    tlsEnabled: true
    collectionName: docs
    vectorSize: 1536
    distance: COSINE

Create a collection for hybrid retrieval, with a named dense vector and a named sparse vector

yaml
id: qdrant_create_hybrid_collection
namespace: company.team

tasks:
  - id: create_collection
    type: io.kestra.plugin.qdrant.CreateCollection
    host: "{{ secret('QDRANT_HOST') }}"
    apiKey: "{{ secret('QDRANT_API_KEY') }}"
    tlsEnabled: true
    collectionName: docs
    vectorName: text_dense
    vectorSize: 4
    distance: COSINE
    sparseVectors:
      - name: text_sparse
        modifier: IDF
Installation

Our standard Docker images already include this plugin. If you use a -slim image or another setup, install it with the following command:

kestractl plugins get io.kestra.plugin:plugin-qdrant --compatible-for $KESTRA_VERSION

For kestractl, use your Kestra version with --compatible-for, or specify an exact plugin version.

Properties

Collection name

The name of the collection to create.

DefaultCOSINE
Possible Values
COSINEDOTEUCLIDMANHATTAN

Distance metric

The distance metric to use for vector comparisons (default: COSINE).

The host of the Qdrant instance

The hostname or IP address of the Qdrant gRPC endpoint.

Vector dimension size

The dimension size of vectors to be stored in the collection.

API key

The API key for authentication with Qdrant.

Assets this task consumes as inputs or produces as outputs, for lineage tracking and the asset graph (Enterprise Edition). A flow declaring this property on a task is rejected in the open-source edition.

Definitions
assetFailureBehaviorstring
Possible Values
IGNOREFAILWARN

Asset failure behavior

Behavior applied to the task state when a declared asset fails to render, emit, or be persisted (e.g. a lock conflict): FAIL escalates it to FAILED, WARN (default) warns it if it would otherwise succeed, IGNORE leaves the state untouched.

enableAutobooleanstring

Whether to auto-register assets referenced dynamically at runtime that are not statically declared in inputs or outputs.

inputsarray

The assets consumed as inputs.

id*string
Min length1
typestring
outputs

The assets produced as outputs.

id*string
Min length1
Max length150
type*object
descriptionstring
displayNamestring
metadataobject
Default{}
namespacestring
Min length1
Max length150
id*string
Min length1
Max length150
type*object
descriptionstring
displayNamestring
metadataobject
Default{}
namespacestring
Min length1
Max length150
id*string
Min length1
Max length150
type*object
descriptionstring
displayNamestring
metadataobject
Default{}
namespacestring
Min length1
Max length150
id*string
Min length1
Max length150
type*object
descriptionstring
displayNamestring
metadataobject
Default{}
namespacestring
Min length1
Max length150
id*string
Min length1
Max length150
type*object
descriptionstring
displayNamestring
metadataobject
Default{}
namespacestring
Min length1
Max length150
id*string
Min length1
Max length150
type*string
Min length1

Custom asset type

descriptionstring
displayNamestring
metadataobject
Default{}
namespacestring
Min length1
Max length150
Defaultfalse

Store vectors on disk

Whether to store vector embeddings on disk (mmap) rather than keeping them entirely in RAM (default: false).

Possible Values
COLDCACHEDPINNED

Payload memory storage policy

Storage memory tier for metadata payload: COLD (on disk/mmap), CACHED (on disk with memory page cache), or PINNED (pinned in RAM). Default: null (server default).

Default6334

The port of the Qdrant instance

The gRPC port of the Qdrant instance (default 6334).

Sparse vector spaces

Named sparse vector spaces to create next to the dense vector. Each name must be unique and different from vectorName. Points get their sparse vectors later through Upsert.

Definitions
name*string

Sparse vector name

Name of the sparse vector space, unique in the collection and different from the dense vector name.

fullScanThresholdintegerstring

Full scan threshold

Number of vectors below which Qdrant prefers a full scan over the sparse index (min: 0). This is a count of vectors, not a size in kilobytes as for the dense HNSW index. Unset keeps the server default.

memoryobject
modifierstring
DefaultNONE
Possible Values
NONEIDF

Weight modifier

NONE (default) uses the stored weights as they are. IDF multiplies them by the inverse document frequency computed by Qdrant, as BM25 does.

Defaultfalse

Enable TLS

Whether to use TLS/SSL for secure gRPC connection.

Dense vector name

Name of the dense vector space. Unset creates the unnamed default vector. Set it when the collection also holds sparse vectors that are queried together with the dense one, for example by HybridQuery.

Collection name

The name of the collection that was created.

Success status

Whether the collection was created successfully.