MemTurbo
A Precog Labs App
Store, search, and retrieve agent memories with industry-leading vector compression. Cut storage costs by 95% while maintaining semantic accuracy.
MemTurbo uses TurboQuant compression to reduce vector storage by up to 95% while preserving semantic search quality. Built for AI agents that need persistent, searchable memory across conversations.
How It Works
Three steps from raw text to searchable compressed memory
Store Memories
Send text content via REST API, Python/TypeScript SDK, or MCP server. MemTurbo ingests it with metadata like agent ID, tags, and session context.
Compress & Embed
TurboQuant generates embeddings and compresses vectors using Lloyd-Max quantization with QJL residual correction for 95% storage reduction.
Search & Retrieve
Query memories with semantic search powered by pgvector HNSW indexing. Results are ranked by cosine similarity with scores, filtered by tags, agent, or session.
Features
Everything you need for AI agent memory management
TurboQuant Compression
Lloyd-Max quantization with QJL residual correction reduces vector storage by up to 95% while maintaining cosine similarity accuracy.
Semantic Search
Find memories by meaning, not just keywords. Cosine similarity search across compressed vectors with relevance scoring via pgvector HNSW.
Multi-Tenant Isolation
Row-level security ensures complete data isolation between organizations and projects. Built for production multi-tenancy from day one.
Agent Memory
Tag memories by agent, user, and session. Build persistent context for AI agents that remember across conversations and interactions.
Version History
Track memory changes over time with automatic versioning. Every update creates a new version, preserving the complete history.
REST API & SDKs
Full REST API with TypeScript and Python SDKs. Store, search, update, and delete memories programmatically with typed clients.
MCP Server
Model Context Protocol server for direct integration with AI assistants like Claude. Expose memories as tools for LLM agents to use natively.
Background Processing
Async embedding pipeline with configurable worker pools. Non-blocking ingestion means your API stays fast while heavy compute runs in background.
API Key Authentication
Simple, secure API key auth with per-key rate limiting. Create multiple keys with different permissions and rate limits per project.
What It Provides
A complete memory infrastructure for AI applications
Persistent storage with metadata, tags, agent IDs, and session tracking
Semantic similarity search with cosine distance scoring via pgvector
TurboQuant 3-bit compression with PolarQuant and QJL residuals
REST API, TypeScript SDK, Python SDK, MCP server, and web dashboard
Simple, Transparent Pricing
Start free, upgrade when you need more storage or collaboration
Ready to add memory to your AI agents?
Get started with the dashboard or integrate directly via the REST API. No credit card required for the free tier.
Launch Dashboard