embeddings

Generate and manage vector embeddings for semantic search, similarity matching, and AI-powered recommendations.

Overview

The embeddings primitive converts text, images, and code into high-dimensional vectors that capture semantic meaning, enabling powerful similarity search and content recommendations.

Parent Primitive: llm - Universal LLM interface

SDK Object Mapping

This primitive maps to the ai SDK object for embedding generation:

import { ai, embeddings } from 'sdk.do'

// AI - Generate embeddings (ai is one of 8 core SDK objects)
const embedding = await ai.embed({
  text: 'Product description...',
  model: 'text-embedding-3-large',
})

// Batch embeddings
const batchEmbeddings = await ai.embedBatch({
  texts: ['Text 1', 'Text 2', 'Text 3'],
  model: 'text-embedding-3-large',
})

// Direct embeddings access
const embedding2 = await embeddings.create({
  text: 'Hello world',
  model: 'text-embedding-3-large',
})

// Search similar content
const similar = await embeddings.search({
  query: 'Hello world',
  limit: 10,
  minScore: 0.7,
})

Quick Example

import { embeddings } from 'sdk.do'

// Generate embeddings
const embedding = await embeddings.create({
  text: 'Hello world',
  model: 'text-embedding-3-large',
})

// Find similar content
const similar = await embeddings.search({
  query: 'Hello world',
  limit: 10,
  minScore: 0.7,
})

Core Capabilities

Text Embeddings - Convert text to semantic vectors
Multi-Modal - Support for text, images, and code
Vector Search - Fast similarity search with cosine similarity
Hybrid Search - Combine semantic and keyword search
Batch Processing - Efficient bulk embedding generation

Access Methods

SDK

TypeScript/JavaScript library for embedding operations

await embeddings.create({ text: 'Product description...', model: 'text-embedding-3-large' })

→ SDK Documentation

CLI

Command-line tool for embedding generation

do embeddings create "Product description..." --model text-embedding-3-large

→ CLI Documentation

API

REST/RPC endpoints for embedding operations

curl -X POST https://api.do/v1/embeddings -d '{"text":"Product description...","model":"text-embedding-3-large"}'

→ API Documentation

MCP

Model Context Protocol for AI-driven embedding operations

Generate an embedding for "Product description..." using text-embedding-3-large model

→ MCP Documentation

Parent Primitive

llm - Universal LLM interface

Sibling Primitives

models - AI model management
vectors - Vector database and search

ai - AI operations (SDK object mapping)
database - Store and query vectors
context - Context-aware embeddings

embeddings

On this page