qdrant-tensr-dev 0.3.1

qdrant-tensr-client — Lightweight Python SDK for the Qdrant Tensr Dev external API

qdrant-tensr-dev

Lightweight Python SDK for the Qdrant Tensr Dev external API.

• Zero required dependencies — uses httpx, requests, or stdlib urllib in that order
• Single file — copy qdrant_tensr_dev.py or install via pip
• Mirrors the Qdrant client API — familiar method names (upsert, search, scroll, …)

Install

pip install qdrant-tensr-dev
# optional: bring your own HTTP library
pip install qdrant-tensr-dev[httpx]
pip install qdrant-tensr-dev[requests]

Quick start

from qdrant_tensr_dev import TensrClient

client = TensrClient(
    base_url="https://your-dashboard-host",
    api_key="YOUR_API_KEY",          # from Dashboard → Groups → API Keys
)

# Create a new collection
col = client.create_collection("my-collection", vector_size=1536, distance="cosine")
# {"id": "...", "name": "my-collection", "vector_size": 1536, "distance": "cosine"}

# List all collections in your group
cols = client.list_collections()
print(cols)
# {"group_id": "...", "collections": [{"name": "...", "vector_size": 384, "distance": "Cosine"}, ...]}

# Get detailed info for one collection
info = client.get_collection("my-collection")

# Upsert vectors
client.upsert("my-collection", points=[
    {
        "id": "550e8400-e29b-41d4-a716-446655440001",
        "vector": [0.1, 0.2, 0.3],
        "payload": {"text": "hello world", "source": "docs"},
    },
])

# Nearest-neighbour search
results = client.search("my-collection", vector=[0.1, 0.2, 0.3], limit=5)
for hit in results:
    print(hit["score"], hit["payload"])

# Cross-collection search across every collection in your group at once
hits = client.search_all(vector=[0.1, 0.2, 0.3], limit=10)
for hit in hits:
    print(hit["collection"], hit["score"], hit["payload"])

# Scroll (paginate) all points
offset = None
while True:
    page = client.scroll("my-collection", limit=100, offset=offset)
    for pt in page["points"]:
        print(pt["id"], pt["payload"])
    offset = page["next_offset"]
    if offset is None:
        break

# Get a single point
pt = client.get_point("my-collection", "550e8400-e29b-41d4-a716-446655440001")

# Count points
total = client.count("my-collection")

# Delete points
client.delete("my-collection", ids=["550e8400-e29b-41d4-a716-446655440001"])

API reference

Method	Description
create_collection(name, vector_size, distance)	Create a new collection and add it to your group
list_collections()	List all active collections in your group
get_collection(name)	Get metadata + live stats for a collection
upsert(collection, points)	Insert or update points
delete(collection, ids)	Delete points by ID
get_point(collection, point_id)	Retrieve a single point
scroll(collection, ...)	Page through all points
count(collection)	Count total points
search(collection, vector, ...)	Nearest-neighbour search (single collection)
search_all(vector, ...)	Cross-collection search across all group collections

All methods raise TensrError(status_code, detail) on HTTP errors.

CLI smoke-test

python -m qdrant_tensr_dev --url https://your-host --key YOUR_KEY