VectorizeAI — API Documentation

🧠 API Reference Documentation

VectorizeAI
Memory Infrastructure

Complete reference for embedding AI knowledge bots, managing vector memory, and building intelligent conversation systems.

Version

v2.0.0

Base URL

https://api.vectorize.ai

Auth

API Key / Session Token

Response

JSON / REST

Table of Contents

1 Authentication 2 Memory Management 3 Conversational AI 4 Data Ingestion 5 Classroom AI Engine 6 Performance Benchmarks 7 Deployment & Embeds 8 Error Reference

1 Authentication

Secure your requests using either a persistent API Key or a temporary Client Session Token. Obtain these from your Dashboard → Developer Settings.

Note: Most endpoints require X-API-Key. Profile and history endpoints require X-Client-Token.

POST/api/clients/loginStandard Authentication

Field	Type	Required	Description
email	string	Required	Registered email address
password	string	Required	Account password

Successful Response

{ 
  "client_id": "clt-...", 
  "token": "clt-...", 
  "name": "Sarah Connor" 
}

2 Memory Management

VectorizeAI uses Memories to silo data. Each memory acts as a dedicated brain for a specific chatbot or department.

POST/api/memoryCreate Knowledge Space

Field	Type	Required	Description
name	string	Required	Unique name for this memory
description	string	Optional	Context for the AI
provider	string	Optional	"gemini" (default) \| "openai"

3 Conversational AI

Query your memories using natural language. The system performs semantic retrieval before generating a grounded response.

POST/api/memory/{id}/askContextual Query

Field	Type	Required	Description
question	string	Required	The user query
top_k	integer	Optional	Chunk depth (default: 5)
history	array	Optional	Conversation thread context

Grounded Response

{
  "answer": "According to your uploaded PDF, the Q4 targets are...",
  "sources": ["Q4_Report.pdf"],
  "latency_ms": 1120
}

4 Data Ingestion

Populate your memory by uploading files or linking dynamic web content. Chunks are automatically vectorized on arrival.

POST/api/memory/{id}/upload-pdfFile Ingestion

Use multipart/form-data with the key file.

POST/api/memory/{id}/ingest-urlWeb Scraping

Field	Type	Description
url	string	Full URL to scrape and index

5 Classroom AI Engine

The Classroom AI Engine exposes a structured hierarchical curriculum system (Exams → Papers → Subjects → Chapters → Topics → Subtopics) along with core RAG educational generation endpoints (Notes, PDF compilation, AI Quiz, and fully synthesized Reels generation).

GET/api/classroom/examsRetrieve All Exams

Fetches the list of all configured exams, including their subjects, chapters, topics, and subtopics.

Response Payload

[
  {
    "exam_id": "exam-...",
    "name": "UPSC Civil Services",
    "papers": [...]
  }
]

POST/api/classroom/papers/{id}/auto-generateAuto-Generate Curriculum Syllabus

Given a syllabus paper, automatically generates chapters, topics, and subtopics using AI, building a complete learning matrix in seconds.

Response

{
  "success": true,
  "chapters_created": 8,
  "topics_created": 32
}

POST/api/classroom/subtopics/{id}/generate-notesCompile RAG Study Notes

Synthesizes extensive, deeply grounded markdown study notes based on RAG knowledge vectors linked to this subtopic.

Response

{
  "success": true,
  "notes": "# Indian Federal Structure\nFederalism is a system of governance..."
}

POST/api/classroom/subtopics/{id}/quiz/generateGenerate AI Custom Quiz

Compiles a custom multiple-choice question quiz from RAG study materials with answers, detailed explanations, and cognitive level mappings.

Response

{
  "success": true,
  "quiz": [
    {
      "question": "Which article of the Constitution describes India as a Union of States?",
      "options": ["Article 1", "Article 2", "Article 3", "Article 4"],
      "answer": "Article 1",
      "explanation": "Article 1 of the Constitution says India, that is Bharat, shall be a Union of States."
    }
  ]
}

POST/api/classroom/subtopics/{id}/generate-reelAuto-Generate Educational AI Reel

Initiates a background job to generate 12 visually photorealistic scene prompts, sync segment-by-segment multilingual narration voiceovers, overlay ASS subtitles, and compile a final mixed video reel with background music.

Field	Type	Required	Description
language	string	Optional	"English" \| "Hindi" \| "Spanish" \| "French" (default: English)
voice_id	string	Optional	Premium ElevenLabs voice personality ID
transcript	string	Optional	Custom lecture text transcript to direct scene scripting

Response

{
  "success": true,
  "job_id": "job-e2bd3664bcc8",
  "scenes_created": 12
}

6 Performance Benchmarks

VectorizeAI is optimized for low-latency voice and text interactions. Typical response times across the pipeline:

Component	Target	Range	Status
Vector Retrieval	~100ms	50–300ms	Fast
Gemini Flash	~800ms	600–1200ms	Optimal
GPT-4o	~1200ms	900–2000ms	Stable
Network (HTTPS)	~150ms	80–400ms	Fast

Optimization: Use streaming for a "first-byte" latency of under 400ms. Perfect for real-time customer support bots.

7 Deployment & Embeds

Drop the VectorizeAI brain into any existing interface with one line of code.

<!-- Instant Iframe Embed -->
<iframe src="https://app.vectorize.ai/embed/{memory_id}" width="100%" height="600"></iframe>