VectorX - Encrypted Vector Database
VectorX is an encrypted vector database designed for maximum security and speed. Utilizing client-side encryption with private keys, VectorX ensures data confidentiality while enabling rapid Approximate Nearest Neighbor (ANN) searches within encrypted datasets. Leveraging a proprietary algorithm, VectorX provides unparalleled performance and security for applications requiring robust vector search capabilities in an encrypted environment.
Key Features
- Client-side Encryption: Vectors are encrypted using private keys before being sent to the server
- Fast ANN Searches: Efficient similarity searches on encrypted vector data
- Multiple Distance Metrics: Support for cosine, L2, and inner product distance metrics
- Metadata Support: Attach and search with metadata and filters
- High Performance: Optimized for speed and efficiency with encrypted data
Installation
pip install vecx
Quick Start
from vecx.vectorx import VectorX
# Initialize client with your API token
vx = VectorX(token="your_user_id:your_api_token:region")
# Generate a secure encryption key
encryption_key = vx.generate_key()
# Create a new index
vx.create_index(
name="my_vectors",
dimension=1536, # Your vector dimension
key=encryption_key, # Encryption key
space_type="cosine" # Distance metric (cosine, l2, ip)
)
# Get index reference
index = vx.get_index(name="my_vectors", key=encryption_key)
# Insert vectors
index.upsert([
{
"id": "doc1",
"vector": [0.1, 0.2, 0.3, ...], # Your vector data
"meta": {"text": "Example document", "category": "reference"}
}
])
# Query similar vectors
results = index.query(
vector=[0.2, 0.3, 0.4, ...], # Query vector
top_k=10,
filter={"category": "reference"} # Optional filter
)
# Process results
for item in results:
print(f"ID: {item['id']}, Similarity: {item['similarity']}")
print(f"Metadata: {item['meta']}")
Basic Usage
Initializing the Client
from vecx.vectorx import VectorX
# Production with specific region
vx = VectorX(token="user_id:api_token:region")
Managing Indexes
# List all indexes
indexes = vx.list_indexes()
# Create an index with custom parameters
vx.create_index(
name="my_custom_index",
dimension=384,
key=encryption_key,
space_type="l2",
M=32, # Graph connectivity parameter
ef_con=200, # Construction-time parameter
use_fp16=True # Use half-precision for storage optimization
)
# Delete an index
vx.delete_index("my_index")
Working with Vectors
# Get index reference
index = vx.get_index(name="my_index", key=encryption_key)
# Insert multiple vectors in a batch
index.upsert([
{
"id": "vec1",
"vector": [...], # Your vector
"meta": {"title": "First document", "tags": ["important"]}
},
{
"id": "vec2",
"vector": [...], # Another vector
"filter": {"visibility": "public"} # Optional filter values
}
])
# Query with custom parameters
results = index.query(
vector=[...], # Query vector
top_k=5, # Number of results to return
filter={"tags": "important"}, # Filter for matching
ef=128, # Runtime parameter for search quality
include_vectors=True # Include vector data in results
)
# Delete vectors
index.delete_vector("vec1")
index.delete_with_filter({"visibility": "public"})
# Get a specific vector
vector = index.get_vector("vec1")
API Reference
VectorX Class
__init__(token=None): Initialize with optional API tokenset_token(token): Set API tokenset_base_url(base_url): Set custom API endpointgenerate_key(): Generate a secure encryption keycreate_index(name, dimension, key, space_type, ...): Create a new indexlist_indexes(): List all indexesdelete_index(name): Delete an indexget_index(name, key): Get reference to an index
Index Class
upsert(input_array): Insert or update vectorsquery(vector, top_k, filter, ef, include_vectors): Search for similar vectorsdelete_vector(id): Delete a vector by IDdelete_with_filter(filter): Delete vectors matching a filterget_vector(id): Get a specific vectordescribe(): Get index statistics and info
Security Considerations
- Key Management: Store your encryption key securely. Loss of the key will result in permanent data loss.
- Client-Side Encryption: All sensitive data is encrypted before transmission.
1 maintainer
Release files for vecx 0.33.3
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| vecx-0.33.3.tar.gz | 4.0 MB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| vecx-0.33.3-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 8.7 MB
Release files / vecx-0.33.3.tar.gz
| Download URL | vecx-0.33.3.tar.gz |
|---|---|
| Size | 4.0 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
0fec1a6757d07f975e3f289b2b75b73923c5f91045c60e9879c89b5f2926ea9e
|
|
BLAKE2b-256 checksum How to use checksums |
7f3c4eb3319c1268f83a7ff4cb36031b9079a83b6ebaa318894eb3b24481495b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.11.11
|
Release files / vecx-0.33.3-py3-none-any.whl
| Download URL | vecx-0.33.3-py3-none-any.whl |
|---|---|
| Size | 4.7 MB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
a790fdfd67f37ba3ec94310822c6d1e858827421a1c5b1a4fcb9c87368cd7d0a
|
|
BLAKE2b-256 checksum How to use checksums |
04cae19477b93abeefff085bcfa3e14f8b6cceb189c195d77c51d168690586eb
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.11.11
|
