Quickstart

Get up and running with BudgetVec in under 5 minutes.

1. Get an API Key

Contact us or use the admin API to create a tenant:

curl -X POST https://budgetvec.standupcode.workers.dev/admin/tenants \
  -H "Content-Type: application/json" \
  -H "X-Admin-Secret: your-admin-secret" \
  -d '{
    "tenant_id": "my-org",
    "display_name": "My Organization",
    "api_key": "bv_live_your_secret_key",
    "plan": "Launch",
    "vector_dim": 768
  }'

2. Install the SDK

npm install @budgetvec/client

3. Initialize the Client

import { BudgetVecClient } from "@budgetvec/client";

const client = new BudgetVecClient({
  baseUrl: "https://budgetvec.standupcode.workers.dev",
  apiKey: process.env.BUDGETVEC_API_KEY!,
});

4. Upsert Vectors

const ns = client.namespace("my-documents");

await ns.upsert({
  upsert_rows: [
    { id: 1, vector: [0.1, 0.2, 0.3, /* ... 768 dims */], title: "Introduction to AI" },
    { id: 2, vector: [0.4, 0.5, 0.6, /* ... 768 dims */], title: "Machine Learning Basics" },
  ],
});

5. Query

Vector Search (ANN)

const results = await ns.query({
  rank_by: ["vector", "ANN", queryVector],
  top_k: 10,
});

Full-Text Search (BM25)

const results = await ns.query({
  rank_by: ["text", "BM25", "search query here"],
  top_k: 10,
});

6. Check Namespace Metadata

const meta = await ns.metadata();
console.log(meta);
// { approx_row_count: 2, dimensions: 768, index: { status: "indexing" }, ... }

Next Steps

• Architecture — understand how BudgetVec works under the hood
• API Reference: Write — full upsert/delete API documentation
• API Reference: Query — query options, filters, and hybrid search
• Hybrid Search — combine vector + BM25 for best results