LanceDB
Serverless, embedded vector database built on Lance columnar format — zero-copy, multimodal, no server needed.
LanceDB
LanceDB is an open-source, serverless vector database built on Lance — a columnar data format optimized for ML workloads. It runs embedded in your application (like SQLite) or as a hosted service, requiring no separate server process. The Lance format enables zero-copy access, fast random reads, and efficient updates — making it ideal for RAG, semantic search, and multimodal AI applications.
Key Features
- Serverless/embedded: no separate server; runs in-process or on S3/GCS/Azure
- Multimodal: store and search vectors alongside text, images, video, audio in one table
- Lance columnar format: 100× faster random access than Parquet for ML patterns
- Full-text search (Tantivy-based) + hybrid search
- Versioning: automatic dataset versioning with time-travel queries
- Python, JavaScript/TypeScript, and Rust APIs
Quick Start
import lancedb
import numpy as np
db = lancedb.connect("./lancedb")
table = db.create_table("embeddings", data=[
{"vector": np.random.rand(384).tolist(), "text": "hello world"},
])
results = table.search(np.random.rand(384).tolist()).limit(5).to_pandas()
Install via ai-supply
npx ai-supply add lancedb-multimodal-vector-store
Curated mirror of the open-source LanceDB (Apache-2.0). Get it from the source.
Compromise signals — malicious or tampered code (leaked secrets, backdoors, a dropped executable) — reduce the score, and known dependency CVEs carry a bounded penalty (they warrant review but never QUARANTINE — update the dependency to clear). Other dangerous-by-capability traits are risk surface, expected for some capabilities. Every finding is mapped to its OWASP control below.
Findings mapped to the OWASP Top 10 for LLM Applications (2025) and the OWASP Machine Learning Security Top 10. Expand any flagged control for the exact findings — compromise reduces the score; expected/risk-surface do not, except a known CVE, which carries a small bounded penalty (high/critical → Review).
The same gate an agent runs before installing (POST /api/v1/trust/lancedb-multimodal-vector-store/check). Click a policy:
Consume LanceDB programmatically. Authenticate with an API key or session — see Authorize an agent.
# Agents: CHECK BEFORE YOU INSTALL (no auth) — score, grade, level, capability manifest
curl https://ai-supply.store/api/v1/trust/lancedb-multimodal-vector-store
# Gate against your org policy (returns { pass, violations })
curl -X POST https://ai-supply.store/api/v1/trust/lancedb-multimodal-vector-store/check \
-H "Content-Type: application/json" \
-d '{"minGrade":"B","denyPermissions":["shell"],"denyUnknownEgress":true}'
# CLI
npx ai-supply add lancedb-multimodal-vector-store
# REST (install → download)
curl -X POST https://ai-supply.store/api/v1/listings/lancedb-multimodal-vector-store/install \
-H "Authorization: Bearer $AIM_KEY"
# MCP tool
install_listing({ "slug": "lancedb-multimodal-vector-store" })OpenAPI spec →Curated mirror — latest upstream source. See the repository for tagged releases.