Serverless Vector Search
VectorDB
Store and search vector embeddings at scale. Build RAG applications, semantic search, and recommendation systems.
RAG Applications
Augment LLMs with your data
Semantic Search
Search by meaning, not keywords
Recommendations
Find similar items instantly
Image Search
Reverse image lookup
Blazing Fast
Sub-millisecond queries at any scale. HNSW indexing for optimal performance.
Auto-scaling
Scales from 100K to billions of vectors automatically. No capacity planning.
Metadata Filtering
Filter results by metadata. Combine vector similarity with structured queries.
Easy Integration
Works with OpenAI, Cohere, BGE, and any embedding model.
Namespaces
Organize vectors by namespace. Multi-tenant by design.
Hybrid Search
Combine vector search with keyword search for best results.
Simple API
Upsert Vectors
import { VectorDB } from '@hostscience/vectordb';
const db = new VectorDB('your-api-key');
// Upsert with OpenAI embeddings
await db.upsert({
namespace: 'docs',
vectors: [
{
id: 'doc-1',
values: await openai.embed(text),
metadata: { title: 'Getting Started', url: '/docs/start' }
}
]
});Query Vectors
// Semantic search
const results = await db.query({
namespace: 'docs',
vector: await openai.embed(userQuery),
topK: 10,
filter: { category: 'tutorials' },
includeMetadata: true
});
// Use results for RAG
const context = results.matches.map(m => m.metadata.content);Pricing
VectorDB Starter
For AI prototypes
Free
- 100,000 vectors
- 100,000 queries/mo
- OpenAI embeddings
- REST API
- Metadata filtering
Most Popular
VectorDB Pro
Production vector search
$49.0/mo
- 1,000,000 vectors
- 1,000,000 queries/mo
- 1M vectors
- Multiple indexes
- Hybrid search
- SDK support
VectorDB Enterprise
Enterprise AI infrastructure
$249.0/mo
- Unlimited vectors
- Unlimited queries/mo
- Unlimited vectors
- Dedicated cluster
- SOC 2
- SLA 99.99%