Official SDKs for Node.js, Python, Go, and React Native. Sub-200ms retrieval, hybrid search, and persistent memory — with just 4 methods.
npm install @basegrid-io/coreimport { BaseGrid } from '@basegrid-io/core';
const bg = new BaseGrid({
apiKey: process.env.BASEGRID_API_KEY
});
// Add a memory
await bg.add({
agentId: 'support-bot',
content: 'User prefers email communication'
});
// Search memories
const results = await bg.search({
agentId: 'support-bot',
query: 'contact preference'
});
console.log(results[0].content);One command to install the native SDK for your language.
Initialize the client with your API key in just one line.
Add and search memories with sub-200ms retrieval times.
Automatically combines semantic vector search with keyword exact-match filtering for the highest accuracy possible.
Blazing fast latency powered by our 3-tier search architecture. P95 latency is strictly guaranteed under 200ms.
Store images alongside text. BaseGrid automatically analyzes images to make visual context searchable semantically.
Built-in AI models automatically scan, categorize, and tag incoming memories so you don't have to manually manage metadata.
Easily manage thousands of AI agents or users within a single API key using isolated namespace IDs.
Framework agnostic API designed to plug directly into your custom code, LangChain workflows, or LlamaIndex pipelines.
Plug BaseGrid directly into LangChain tools and persistent memory components.
View Example