Give your AI persistent, searchable memory. Sub-200ms semantic search with multimodal embeddings. Free to start.
Give your ClawdBot, MoltBot, and custom OpenClaw agents the memory they need to truly understand and assist.
OpenClaw agents are powerful, but without memory they start fresh every session. Neutron gives them persistent, searchable memory - so they remember user preferences, past conversations, and learned context across every interaction.
Without memory, AI agents forget everything between sessions. With Neutron, your agents remember, learn, and get smarter over time.
Agents recall past conversations, user preferences, and previous decisions without re-asking.
Build knowledge from every interaction. Your agents become more helpful the more they're used.
No more repeating information. Agents retrieve what they need instantly via semantic search.
Built for speed, scale, and developer happiness.
Lightning-fast semantic search powered by PostgreSQL with pgvector. Your agents get instant recall.
1024-dimensional Jina v4 embeddings. Search across text, images, and documents in a unified space.
Built-in support for external users. Each of your users gets isolated, secure memory spaces.
Organize knowledge into logical groups. Perfect for projects, topics, or conversation threads.
Clean REST API with a TypeScript SDK. Integrate in minutes, not days.
Jina v4 handles 100+ languages natively. Build global applications without translation overhead.
From personal assistants to enterprise knowledge bases.
Build AI that remembers preferences, past conversations, and learns from every interaction.
Ground your LLM responses in your own data. Reduce hallucinations with semantic retrieval.
Context-aware support that remembers customer history and past resolutions.
Index documents, wikis, and notes. Find anything with natural language queries.
Power ClawdBot, MoltBot, and other OpenClaw ecosystem agents with persistent memory.
Each user gets their own isolated memory space. Perfect for SaaS AI products.
Simple API, powerful results. Here's all you need to get started.
// Agent remembers user's preferences from conversation
const storeMemory = await fetch('https://api.neutron.vanarchain.com/v1/seeds', {
method: 'POST',
headers: {
'Authorization': 'Bearer nk_...',
'Content-Type': 'application/json'
},
body: JSON.stringify({
content: "Mike's usual coffee order is an oat milk latte with an extra shot, no sugar. He goes to Blue Bottle on weekday mornings before work.",
metadata: { userId: "user_mike", type: "preference" }
})
});
// Next morning: Agent recalls when Mike asks for help
const recallMemories = await fetch('https://api.neutron.vanarchain.com/v1/seeds/search', {
method: 'POST',
headers: {
'Authorization': 'Bearer nk_...',
'Content-Type': 'application/json'
},
body: JSON.stringify({
query: "What's my usual coffee order?",
limit: 5
})
});Join developers building smarter AI applications with Neutron.
Get Started Free