A lightweight, optimized vector database built on Val Town's SQLite (powered by Turso/libSQL) with Nebius Qwen3-Embedding-8B embeddings.

Metric	Value
Embedding dimensions	4096
Storage per record	~22 KB
Max records per 1GB	~47,500
Avg embedding time	~460ms
Recommended maxDistance	0.6 - 0.65

• Semantic search with cosine similarity + distance scores
• Smart re-embedding - only re-embeds when text changes
• Optimized index - 75% less storage with float8 compression
• Scale testing tools - seed data, calibrate thresholds, detailed stats

Method	Endpoint	Description
POST	/upsert	Insert/update {id, text, meta?}
POST	/search	Search {query, k?, maxDistance?}
POST	/delete	Delete {id}
GET	/get?id=...	Get single record
GET	/list?limit=...	List record IDs

Method	Endpoint	Description
GET	/ping	Health check
GET	/stats	Detailed storage stats
GET	/seed?n=100	Seed N synthetic records
GET	/calibrate?q=...	Suggest distance thresholds
POST	/reindex	Recreate optimized index
POST	/clear?confirm=yes	Delete ALL records

Distance	Meaning	Action
0.0 - 0.4	Very similar	Always include
0.4 - 0.6	Related	Include (tight mode)
0.6 - 0.7	Somewhat related	Include (balanced mode)
0.7+	Likely unrelated	Filter out

Default recommendation: maxDistance: 0.64

import * as db from "https://esm.town/v/kamenxrider/slimarmor/vectordb.ts";

await db.setup();
await db.upsert("doc-1", "Your text here", { category: "notes" });
const results = await db.search("search query", 10, 0.64);

Variable	Description
NEBIUS_API_KEY	Nebius API key for embeddings

• Runtime: Val Town (Deno)
• Database: Val Town SQLite (Turso/libSQL with DiskANN)
• Embeddings: Nebius Qwen3-Embedding-8B (4096 dims)
• Index: Cosine similarity, max_neighbors=64, compress_neighbors=float8

MIT