in progress

Python sdk

two python clients ship with the project, pick the one that fits your workflow.

which sdk should i use?

sdk	connects to	requires	use case
python_exporter	go exporter	exporter service running	prometheus-style monitoring
qdrant_cloud_clientin progress	qdrant api directly	qdrant access	independent / cloud monitoring

exporter sdk

this sdk reads the go exporter's /metrics endpoint. run the go exporter first.

install

bash

python -m pip install -r requirements.txt
python -m pip install -e .

usage

example.py

from python_exporter import ExporterClient

client = ExporterClient("http://localhost:9999")

# list every metric name
print(client.metric_names())

# read a single value
print(client.get_value("qdrant_up"))

# read every sample for a labeled metric
print(client.get_samples("qdrant_collection_points"))

# pick a sample by label
print(client.get_value(
    "qdrant_collection_points",
    {"collection": "my_collection"},
))

# dump all metrics as json
print(client.to_json())

what you get

metric names pulled live from the exporter
values for any metric the exporter exposes
labeled collection data with no hardcoding
automatic support for new metrics, the sdk stays in sync with go

live test

if the exporter is already running, run the integration suite:

bash

export EXPORTER_URL=http://localhost:9999
python -m pytest tests/test_live_exporter_sdk.py -m integration

or inspect a few collections without pytest:

bash

export EXPORTER_URL=http://localhost:9999
python tests/inspect_exporter.py

qdrant cloud sdkin progress

this one talks directly to the qdrant api. no go exporter required.

install

bash

python -m pip install -r requirements_cloud_sdk.txt
python -m pip install -e .

usage

cloud_example.py

from qdrant_cloud_client import QdrantCloudClient

# local qdrant
client = QdrantCloudClient(base_url="http://localhost:6333")

# or qdrant cloud
client = QdrantCloudClient(
    base_url="https://your-cluster.qdrant.io",
    api_key="your-api-key",
)

# list collections
collections = client.get_collections()

# detailed info
info = client.get_collection_info("my_collection")
print(f"vectors: {info.vectors_count}")

# prometheus-style metrics
metrics = client.get_metrics()
for coll in metrics:
    for name, value in coll.items():
        print(f"{name} {value}")

# render as prometheus text
print(client.format_prometheus_metrics())

what you get

direct connection to qdrant cloud
collection info without the go exporter
same metric shape as the exporter
local and cloud support
api-key authentication
prometheus-compatible text output

testing

bash

python -m pytest tests/test_qdrant_cloud_client.py -v

need a full example?

check examples/qdrant_cloud_usage.py in the repo, it covers health checks, error handling, and metric extraction.