Everything runs on your machine.
PRODUCT
RESEARCH50+ MCP tools, one install, zero cloud.
Conversation turns, tool outputs, observations any unstructured agent context fed in as text.
Every memory evaluated: ADD, UPDATE, DELETE, or NO_OP if already known. Zero duplicates.
Persisted across 4 graph types in SQLite. Survives all session resets. REM compresses while idle.
Spec-decoding retrieval bi-encoder shortlist re-ranked by cross-encoder. Ranked, scored, graph-aware.
Every new input is evaluated: ADD, UPDATE, DELETE, or NO_OP. Zero drift. Zero bloat.
Similarity between memories. Finds related concepts across your full context history.
Cause → effect relationships. Understands why things happened, not just what.
Before → after sequences. Tracks how knowledge evolves and decays over time.
Named entity co-occurrence. Connects people, projects, and events automatically.
// 1. Install
// npm install vektor-slipstream
import { createMemory } from 'vektor-slipstream';
// 2. Initialise
const memory = await createMemory({
provider: 'gemini',
apiKey: process.env.GEMINI_API_KEY,
agentId: 'my-agent',
dbPath: './my-agent.db',
});
// 3. Remember
await memory.remember("User prefers TypeScript");
// 4. Recall
const ctx = await memory.recall("coding preferences");Pure SQLite. No cloud dependency. Your memory graph never leaves your server.
Claude, Gemini, Groq, Mistral, OpenAI, Ollama, OpenRouter switch with one config change.
Automatic curation prevents contradictions and duplicates. No manual management.
Compresses 50 fragments into 3 core insights while your agent is idle.
50+ MCP tools, one install, zero cloud.
VEKTOR is an active memory layer it evolves, curates, and reasons. Never passive.
Standard RAG vector stores
VEKTOR Memory
The GUI, CLI and 50+ MCP-DXT tools are all inclusive across 24 integrated connectors.
LangChainDrop-in memory layer for LangChain agents.
recall() returns context, remember() stores.
v1 + v2 adapters included.
OpenAI Agents SDKPersistent memory for OpenAI agent loops.
Recalled context injected into system prompt.
GPT-4o and o-series models supported.
Claude MCP ServerFull MCP module vektor_recall, vektor_store,
cloak_ssh_exec, cloak_fetch tools.
Connect Claude Desktop in minutes.
Gemini /Provider-agnostic single config switch.
Key pooling for Gemini multiple API keys,
waterfall rotation, zero rate-limit downtime.
Mistral MCPvektor_memoire HTTP tool for Le Chat
and Mistral API agents. Local bridge on
localhost:3847. French-first sovereign memory.
CLOAK50+ tool MCP layer for Claude Desktop.
Stealth browser, credential vault, CAPTCHA solving,
behaviour injection. Zero cloud. One install.
Built by a 30-year IT, storage & data networks veteran PET privacy-enhancing technology, our customers’ privacy is a focus Security & compliance docs are public, not behind a form
Cancel anytime, one click Your memories are your SQLite file keep and migrate them, always, no lock-in, ever Every question answered up front, no sales call needed
Benchmark results published LongMemEval, peer-reviewed SOC-style disclosure available on request Open-source core, five Apache 2.0 tools, audit it yourself
No cloud. No embedding bill. No data handshake. VEKTOR runs on your machine, under your control, private backed by five Apache 2.0 CLI tools.
Free, open-source companions each one solves a piece of the memory-infrastructure puzzle so VEKTOR doesn’t have to.
No cloud, no add-ons, no upsell ladder every core system VEKTOR ships with, bundled into one flat subscription.