Skip to main content

Quick Start Guide

Create a Qdrant vector database, connect to it, and run your first similarity search.

Prerequisites​

  • TIR Account: An active TIR project with sufficient credits.
  • Python 3.8+: (Optional) Required if you plan to use the Python client. You may use TIR Nodes for easy execution.

Step 1: Navigate to Vector Database​

  1. Log in to the TIR Dashboard.
  2. Create or select a Project.
  3. Click Vector Database in the sidebar.
  4. Click CREATE DATABASE.

Step 2: Configure Your Database​

Choose a Plan​

Select the plan that matches your workload. The plan determines the CPU, RAM, and GPU resources available per node.

SettingDescription
Cluster TypeTIR Cluster (managed) or Private Cluster (your own infrastructure)
PlanCPU/RAM/GPU allocation per node
Node Count3 to 10 nodes (minimum 3 for fault tolerance)
Disk Size10 to 1,000 GB per node
Disk EncryptionAES-256 encryption at rest (irreversible once enabled)
tip

Clusters with 3 or more nodes can tolerate one node failure. The higher the replication factor, the more resilient your database.

Configure Default Collection​

SettingDescription
Collection NameName for your initial collection
Vector DimensionsSize of your vectors (1 to 4,096)
Distance MetricCosine, Euclid, Dot, or Manhattan
Shard NumberNumber of data partitions (set as a multiple of node count)
Replication FactorNumber of shard copies across nodes (minimum 3)

The bottom panel displays the approximate vector capacity and pricing based on your configuration.

Step 3: Launch​

Review your cluster configuration and cost summary on the bottom right, then click Launch.

info

The cluster takes a few minutes to provision. Status changes from Creating to Running.

Step 4: Get Your Connection Details​

Once the database is Running, click on it to open the Overview tab.

DetailDescription
Endpoint URLYour cluster's hostname (copyable)
API KeyFull read/write access key (keep secure)
Read-Only API KeyRead-only access key for query-only clients
PortProtocol
6333REST (HTTP)
6334gRPC
danger

Keep your API keys secure. Do not share them or commit them to version control.

Step 5: Connect and Create a Collection​

Using Python Client​

pip install qdrant-client
from qdrant_client import QdrantClient
from qdrant_client.http.models import Distance, VectorParams, PointStruct

client = QdrantClient(
    host="<your-endpoint-url>",
    port=6333,
    api_key="<your-api-key>"
)

# Create a collection
client.create_collection(
    collection_name="test_collection",
    vectors_config=VectorParams(size=4, distance=Distance.DOT),
    shard_number=6,
    replication_factor=2
)
info

For gRPC connections, use port 6334 and add prefer_grpc=True to the client constructor.

Using HTTP (cURL)​

curl -X PUT \
  -H "api-key: <your-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "vectors": { "size": 4, "distance": "Dot" },
    "shard_number": 6,
    "replication_factor": 2
  }' \
  https://<your-endpoint-url>:6333/collections/test_collection

Step 6: Insert and Search Vectors​

Insert Points​

points = [
    PointStruct(id=1, vector=[0.05, 0.61, 0.76, 0.74], payload={"city": "Berlin"}),
    PointStruct(id=2, vector=[0.19, 0.81, 0.75, 0.11], payload={"city": "London"}),
    PointStruct(id=3, vector=[0.36, 0.55, 0.47, 0.94], payload={"city": "Moscow"}),
    PointStruct(id=4, vector=[0.18, 0.01, 0.85, 0.80], payload={"city": "New York"}),
    PointStruct(id=5, vector=[0.24, 0.18, 0.22, 0.44], payload={"city": "Beijing"}),
    PointStruct(id=6, vector=[0.35, 0.08, 0.11, 0.44], payload={"city": "Mumbai"}),
]

client.upsert(collection_name="test_collection", points=points, wait=True)

Search for Similar Vectors​

results = client.search(
    collection_name="test_collection",
    query_vector=[0.2, 0.1, 0.9, 0.7],
    limit=3
)

for point in results:
    print(f"ID: {point.id}, City: {point.payload['city']}, Score: {point.score}")

Manage Your Database​

Start and Stop​

  1. Click the Actions icon (arrow) for your database.
  2. Select Stop to pause or Start to resume.
info

Stopping pauses compute billing, but storage charges continue as long as the database exists.

Delete​

  1. Click the Actions icon and select Delete.
  2. Confirm the deletion.
Permanent Action

Deleting a database permanently removes all data, collections, and configurations. Take a snapshot before deletion if you need to preserve data.

Next Steps​