Production-ready vector infrastructure

AI-Native Vector Database

Built for developers who need semantic search, RAG, and similarity matching at scale. Zero configuration required.

Start Building View Benchmarks
10B+
Vectors indexed
<50ms
P95 latency
99.9%
Uptime SLA
from vectordb import Client

# Initialize client
client = Client("your-api-key")

# Create collection
collection = client.create_collection(
    name="embeddings",
    dimension=1536
)

# Insert vectors
collection.insert([
    {"id": "doc_1", "vector": embedding, "metadata": {...}},
    {"id": "doc_2", "vector": embedding, "metadata": {...}}
])

# Semantic search
results = collection.search(
    vector=query_embedding,
    limit=10,
    filter={"category": "technology"}
)

for result in results:
    print(f"Score: {result.score}, ID: {result.id}")

Built for Production

Enterprise-grade vector infrastructure designed for ML teams who ship.

Vector Search

HNSW and IVF algorithms with approximate nearest neighbor search. Configurable recall/speed tradeoffs.

Hybrid Queries

Combine vector similarity with metadata filters and full-text search. Single unified query language.

Real-Time Indexing

Write vectors and query immediately. No batch processing delays. Sub-second index updates.

Horizontal Scaling

Auto-sharding across nodes. Scale to billions of vectors without reshaping your infrastructure.

Designed for Scale

Architecture that handles production workloads out of the box.

Query Latency (P95) <50ms
Index Build Speed 1M/sec
Recall @ 10 98.5%
Concurrent Queries 10K+
Distributed architecture visualization

Distributed Architecture

Raft consensus for metadata coordination. Sharded indexes across storage nodes. Automatic replication and failover with zero downtime deploys.

Use Cases

Production-tested across semantic search, RAG, and recommendation systems.

RAG applications

RAG Pipelines

Build retrieval-augmented generation systems with context-aware LLM responses. Store document embeddings, retrieve relevant context, and feed to your language model.

  • Document Q&A systems
  • Knowledge base chat
  • Contextual assistants
Semantic search

Semantic Search

Move beyond keyword matching. Search by meaning, not just exact text. Handle typos, synonyms, and multilingual queries automatically.

  • E-commerce product search
  • Code similarity detection
  • Research paper retrieval
Recommendation systems

Recommendations

Power content and product recommendations with vector similarity. Find similar items, users, or patterns in high-dimensional embedding spaces.

  • Content discovery
  • Similar item matching
  • Personalization engines

Transparent Pricing

Start free, scale when you're ready. No hidden fees.

Free

$0 /month
  • 1M vectors
  • 100K queries/month
  • Community support
  • Single region
POPULAR

Pro

$99 /month
  • 100M vectors
  • 10M queries/month
  • Priority support
  • Multi-region
  • 99.9% SLA

Enterprise

Custom
  • Unlimited vectors
  • Unlimited queries
  • 24/7 dedicated support
  • Private cloud deployment
  • Custom SLA

Quick Start

From zero to production in under 5 minutes.

1

Install the SDK

# Python
pip install vectordb-python

# JavaScript/TypeScript
npm install @vectordb/client
2

Get API Key

Sign up at vectordb.ai/signup to get your API key. Free tier includes 1M vectors and 100K queries/month.

3

Start Querying

Create collections, insert vectors, and start searching. Check the full API reference for advanced filtering and hybrid queries.

example.py 
import vectordb
from openai import OpenAI

# Initialize clients
vdb = vectordb.Client(api_key="your-key")
openai = OpenAI()

# Create collection
collection = vdb.create_collection(
    name="documents",
    dimension=1536,
    metric="cosine"
)

# Generate embeddings and insert
texts = ["AI is transforming software", ...]
embeddings = openai.embeddings.create(
    model="text-embedding-3-small",
    input=texts
)

for i, (text, emb) in enumerate(zip(texts, embeddings.data)):
    collection.insert({
        "id": str(i),
        "vector": emb.embedding,
        "metadata": {"text": text}
    })

# Query with filters
query = "machine learning applications"
query_emb = openai.embeddings.create(
    model="text-embedding-3-small",
    input=query
).data[0].embedding

results = collection.search(
    vector=query_emb,
    limit=5,
    filter={"category": "ai"}
)

for result in results:
    print(f"{result.score:.3f}: {result.metadata['text']}")

API Examples

Python and JavaScript SDK with identical APIs.

Python SDK

# Hybrid search: vector + metadata filter
results = collection.search(
    vector=embedding,
    limit=10,
    filter={
        "category": {"$in": ["tech", "ai"]},
        "published_at": {"$gte": "2024-01-01"}
    }
)

# Batch upsert with conflict resolution
collection.upsert(
    vectors=[...],
    on_conflict="update"
)

# Aggregations with vector search
stats = collection.aggregate(
    vector=embedding,
    group_by="category",
    metrics=["count", "avg_score"]
)

JavaScript SDK

// Hybrid search: vector + metadata filter
const results = await collection.search({
    vector: embedding,
    limit: 10,
    filter: {
        category: { $in: ['tech', 'ai'] },
        published_at: { $gte: '2024-01-01' }
    }
});

// Batch upsert with conflict resolution
await collection.upsert({
    vectors: [...],
    onConflict: 'update'
});

// Aggregations with vector search
const stats = await collection.aggregate({
    vector: embedding,
    groupBy: 'category',
    metrics: ['count', 'avg_score']
});
View Full API Reference

Get in Touch

Enterprise inquiries or technical questions? We're here to help.

Enterprise Sales

Discord Community

Join 5,000+ developers building with VectorDB. Get help, share projects, and connect with the team.

Join Discord →

GitHub

Explore examples, SDKs, and integrations. Star the repo to stay updated on releases.

View on GitHub →

Documentation

Complete guides, API reference, and tutorials to get you from prototype to production.

Read Docs →
made by CoolPages