All Tools
S
Vector DBFreemium
SCYLLADB
High-performance NoSQL database with native vector search
ScyllaDB Source Available License
ABOUT
Traditional NoSQL databases lack native vector embedding support for AI workloads, while dedicated vector databases struggle with high-throughput transactional workloads, forcing teams to maintain separate systems. ScyllaDB unifies a high-performance wide-column store with native HNSW-based ANN vector search, enabling real-time AI applications — RAG, semantic caching, feature stores — at billion-scale with sub-2ms P99 latency within a single Cassandra-compatible database.
INSTALL
docker pull scylladb/scylla
INTEGRATION GUIDE
1. Store and retrieve document and embedding vectors for retrieval-augmented generation to augment LLM responses with domain-specific context at scale
2. Cache LLM responses based on embedding similarity to reduce API costs and improve response latency for repeated or similar queries
3. Serve millions of feature vectors per second for online model inference in recommendation, fraud detection, and personalization systems
4. Run billion-scale product, image, or video semantic similarity search using ANN (HNSW) with sub-2ms P99 latency
TAGS
nosqlwide-columncassandra-compatibledynamodb-compatiblevector-databaseannhnswsimilarity-searchragsemantic-cachingdistributed