Build smarter AI with persistent memory

The cost-effective API that gives your AI agents long-term memory. With local embeddings and transparent pricing, store context at scale without breaking the bank.

~200ms
Avg response time
99.9%
Uptime SLA (Pro+)
$0
Embedding costs
GDPR
Compliant

Instant setup.
No config.

Get started in seconds with our simple REST API. No complex setup, no infrastructure to manage.

terminal
$ npm install @persistq/sdk

# Ready to use in 2 seconds

Zero embedding costs.
Save thousands.

Unlike competitors who charge for OpenAI embeddings, we generate vectors locally using open-source models. No per-token fees. No surprise bills.

  • Free embeddings forever
    Transformers.js runs locally - no API costs
  • Privacy-first
    Your data never leaves your infrastructure for embeddings
  • Transparent pricing
    Flat rates based on storage, not usage

Cost Comparison

Competitor (OpenAI embeddings)
1M memories × $0.0001/embedding = $100+/month
PersistQ (local embeddings)
1M memories = $0 embedding costs
Save $100+/month
on every million memories
Claude Code
> One prompt, instant setup
"Add PersistQ memory to my AI agent using MCP"
✓ MCP server configured
✓ Memory tools available
✓ Ready to use in seconds

Works with Claude Code.
MCP ready.

One prompt and Claude Code sets up PersistQ via MCP (Model Context Protocol). No manual configuration needed.

  • MCP integration
    Automatic setup with Claude Code
  • Zero config
    Claude handles all the setup for you

Works with your stack

Integrate PersistQ into any language or framework within minutes

import { PersistQ } from '@persistq/sdk';

const memory = new PersistQ(process.env.MEMORY_API_KEY);

// Store a memory
await memory.store({
  content: 'User prefers dark mode',
  group: 'preferences',
  tags: ['ui', 'settings']
});

// Retrieve memories
const memories = await memory.search({
  query: 'user preferences',
  group: 'preferences'
});

Semantic search.
Built-in.

Vector-based search finds relevant memories even with fuzzy queries. No need to remember exact phrases.

  • Automatic embeddings
    We handle vector generation for you
  • Hybrid search
    Combines semantic and keyword search
  • Zero embedding costs
    Local Transformers.js - no per-token charges
Query: "user color preferences"
Finds: "User prefers dark mode"
Query: "what does the user like"
Finds: "User enjoys sci-fi movies"
North America12ms
Europe18ms
Asia Pacific24ms
South America31ms

Lightning fast.
Built for scale.

PostgreSQL with pgvector extension ensures sub-200ms semantic search across millions of memories. Optimized for real-time applications.

Security first.
Privacy built-in.

GDPR compliant with secure authentication. Your data stays private with local embedding generation.

Encrypted storage
GDPR compliant
Secure API key authentication
Local embedding generation
GDPR compliant data handling
User data deletion on request
No third-party AI dependencies

Simple, transparent pricing

Start free. Scale as you grow. No hidden fees.

Free

$0
  • 5,000 API calls/mo
  • 250 memories
  • 12.5MB storage

Starter

$5/mo
  • 50K API calls/mo
  • 2,500 memories
  • 250MB storage
Popular

Pro

$12/mo
  • 500K API calls/mo
  • 25,000 memories
  • 5GB storage

Premium

$29/mo
  • 2M API calls/mo
  • 100K memories
  • 50GB storage

Ready to build smarter AI?

Start building with PersistQ today. No credit card required.