RAG as a Service

Semantic search for AI agents. Index your documents, search with natural language. Simple API, powerful results.

View Docs
# Index a document curl -X POST https://vectorpass-api.automators.work/v1/index \ -H "X-API-Key: vp_live_xxx" \ -d '{"id": "doc1", "text": "Your document text"}' # Search semantically curl -X POST https://vectorpass-api.automators.work/v1/search \ -d '{"query": "find relevant content", "k": 10}'

Everything you need for semantic search

Lightning Fast

Sub-millisecond search latency. Scales to millions of vectors effortlessly.

Multilingual

Search in 100+ languages out of the box. No configuration needed.

Secure

API key authentication, rate limiting, and data isolation per account.

Hybrid Search

Combine semantic vectors with BM25 keyword search for best results.

No Infrastructure

Runs on Cloudflare edge. No servers to manage, scales automatically.

Simple API

RESTful API with just 5 endpoints. Integrate in minutes, not days.

Simple, transparent pricing

Start free, scale as you grow

Free

$0 /month
  • 1,000 vectors
  • 100 searches/day
  • 1 database
  • Batch size: 10

Starter

$9 /month
  • 50,000 vectors
  • 10K searches/day
  • 5 databases
  • Batch size: 100
Popular

Pro

$29 /month
  • 500,000 vectors
  • 100K searches/day
  • 20 databases
  • Batch size: 500

Business

$79 /month
  • 5M vectors
  • 1M searches/day
  • 100 databases
  • Batch size: 1,000
Referral Program

Earn up to 50% off your subscription

Share your referral code with friends. For each friend with an active subscription, you get 10% off. Stack up to 5 active referrals for 50% off.

10%
per active referral
50%
max discount
Ongoing
while they pay

Documentation

Everything you need to integrate semantic search into your AI agents

1 Get your API key

Register for free and verify your email. Your API key will be displayed immediately.

2 Index documents

curl -X POST https://vectorpass-api.automators.work/v1/index \ -H "X-API-Key: YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "id": "doc1", "text": "Your document content here. Can be any length!", "metadata": {"category": "docs", "author": "John"} }'
Automatic Chunking: Long documents are automatically split into ~1500 character chunks with 200 char overlap. Each chunk is indexed separately for better retrieval.

3 Search semantically

curl -X POST https://vectorpass-api.automators.work/v1/search \ -H "X-API-Key: YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{"query": "find relevant documents", "k": 10}'

Response:

{ "success": true, "data": { "results": [ { "id": "doc1", "distance": 0.234, "metadata": {"category": "docs", "_snippet": "Your document..."} } ], "query": "find relevant documents", "k": 10 } }

MCP Server

Model Context Protocol for AI Assistants

Connect VectorPass directly to Claude Desktop, Cursor, or any MCP-compatible AI client. Use natural language to index documents and search your knowledge base.

MCP Server URL

https://vectorpass-mcp.automators.work

Authentication

OAuth 2.1 with your VectorPass email. No API key needed - just login with your email and verify with a code.

Available Tools

vectorpass_index

Index a document

vectorpass_search

Semantic search

vectorpass_keyword_search

BM25 keyword search

vectorpass_delete

Delete document

vectorpass_batch_index

Index multiple docs

vectorpass_list_databases

List all databases

vectorpass_create_database

Create new database

vectorpass_stats

Usage statistics

Connect from Claude Desktop

Add this to your Claude Desktop MCP settings:

{ "mcpServers": { "vectorpass": { "url": "https://vectorpass-mcp.automators.work" } } }

On first use, you'll be prompted to login with your VectorPass email. A verification code will be sent to authorize the connection.

API Reference

Database Management

Endpoint Description Body
GET /v1/databases List all databases -
POST /v1/databases Create new database {name}
GET /v1/databases/:name Get database info -
DELETE /v1/databases/:name Delete database -

Vector Operations

Endpoint Description Body
POST /v1/index Index a document {id, text, metadata?, db?}
POST /v1/batch Batch index (up to tier limit) {items: [{id, text, metadata?}], db?}
POST /v1/search Semantic vector search {query, k?, db?}
POST /v1/keyword BM25 keyword search {query, k?, db?}
DELETE /v1/vectors/:id Delete document + chunks ?db=name
GET /v1/stats Usage statistics ?db=name
Multi-Database: All vector endpoints accept an optional db parameter. If not specified, defaults to "default". Database limits by tier: Free (1), Starter (5), Pro (20), Business (100).

Multi-Database Support

  • Create multiple isolated databases
  • Organize data by project, tenant, or use case
  • Vector limits shared across all databases
  • Default database created automatically

Automatic Chunking

  • Long documents split into ~1500 char chunks
  • 200 char overlap for context continuity
  • Intelligent sentence boundary detection
  • Each chunk indexed as id:chunk_N

Semantic Search

  • 100+ languages supported
  • Understands meaning, not just keywords
  • Sub-100ms search latency
  • BM25 keyword search included

Authentication

  • Header: X-API-Key: vp_live_xxx
  • Or: Authorization: Bearer vp_live_xxx
  • Email verification required for writes
  • Rate limits per tier (see Pricing)

Response Format

  • All responses: {success, data?, error?}
  • Search results include distance (0-2)
  • Lower distance = more similar
  • Metadata preserved from indexing

Database Limits by Tier

  • Free: 1 database
  • Starter: 5 databases
  • Pro: 20 databases
  • Business: 100 databases