Introducing DotVectorDB: Serverless Vector Database for AI Applications

Why We Built DotVectorDB

Every AI application needs two things: powerful models and relevant context. While we already provide access to 100+ AI models through our gateway, we saw developers struggling with the context part—building semantic search, managing embeddings, and implementing RAG (Retrieval Augmented Generation).

Existing vector databases are either too complex, too expensive, or require managing infrastructure. We wanted something different: a serverless vector database that just works.

What Makes It Different

DotVectorDB vs Turbopuffer

Both are serverless, but DotVectorDB is built for knowledge graphs. While Turbopuffer does vector search, DotVectorDB supports KGE (Knowledge Graph Embeddings) for relationship traversal and complex queries.

Get Started in Minutes

Here's how simple it is to build a RAG application:

curl -X POST https://api.infraxa.ai/api/vector/tenants \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"tenant_id": "my-app"}'

curl -X POST https://api.infraxa.ai/api/vector/tenants/my-app/blobs \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "blob_id": "docs",
    "blob_type": "dense",
    "dimension": 768
  }'

// Generate embeddings with Gemini
const embedding = await generateEmbedding(text);

// Insert into DotVectorDB
await fetch('https://api.infraxa.ai/api/vector/tenants/my-app/blobs/docs/vectors', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${API_KEY}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    vectors: [embedding],
    attributes: [{ text, title, url }],
  }),
});

// Search for context
const results = await searchVectors(query);
const context = results.attributes.map(a => a.text).join('\n\n');

// Generate AI response with context
const response = await streamText({
  model: infraxa("z-ai/glm-4.5-flash"), // FREE!
  messages: [{ role: "user", content: query }],
  system: `Use this context:\n\n${context}`,
});

Build AI Agents That Can Pay for Data

DotVectorDB + x402 enables truly autonomous AI agents. Your agents can search premium knowledge bases, access paid APIs, and retrieve specialized data—all without human intervention.

import { Agent } from "agents";
import { withX402Client } from "agents/x402";

export class ResearchAgent extends Agent {
  async searchPremiumKnowledge(query: string) {
    // Build x402 client with agent's wallet
    const x402Client = withX402Client(
      this.mcpClient,
      { network: "base", account: this.account }
    );

    // Agent automatically pays to access premium vector DB
    const results = await x402Client.callTool(
      null, // Auto-approve payments
      {
        name: "search_vectors",
        arguments: { 
          query,
          database: "premium-research-papers"
        }
      }
    );

    // Agent now has access to premium research data
    return results;
  }
}

Transparent Pricing

Start free, scale as you grow. No hidden fees, no surprises.

Alpha Release & Roadmap

DotVectorDB is currently in alpha. We're actively developing new features and would love your feedback.

Join our alpha program and help shape the future of DotVectorDB. Your feedback directly influences our roadmap.