Documentation
Everything you need to build AI agents with instant, global memory. From quick start to advanced patterns.
Get Started
Quick Example
Agent Memory in 30 seconds
import { VectorsDB } from '@aetherfy/vectorsdb'
const client = new VectorsDB({
apiKey: process.env.VECTORSDB_API_KEY
})
// Create collection for agent memories
await client.createCollection('agent-memory', {
vector_size: 1536,
distance: 'cosine'
})
// Store agent memory
await client.upsert('agent-memory', {
id: 'memory-001',
vector: embeddings,
metadata: {
agent_id: 'assistant',
content: 'User prefers morning meetings',
timestamp: Date.now()
}
})
// Search memories
const results = await client.search('agent-memory', {
vector: queryEmbedding,
limit: 5,
filter: { agent_id: 'assistant' }
}) // Returns in <100ms globally
Live Performance
23ms
US East
41ms
EU West
67ms
Asia Pacific
Built for AI Agents
Global < 100ms
Intelligent edge caching ensures your agents have instant access to memories worldwide
Developer First
Familiar APIs, comprehensive docs, and migration tools that just work
$
Pay Per Use
Transparent per-operation pricing. No pod management or surprise bills