All Tools
D
Vector DBFreeOpen Source
DRAGONFLY
High-throughput Redis-compatible data store with vector search
BSL-1.1
ABOUT
Traditional Redis deployments struggle with single-threaded throughput limitations as AI workloads scale — especially for vector similarity search, LLM response caching, and real-time feature serving. Dragonfly solves this with a multi-threaded, shared-nothing architecture that delivers up to 25x higher throughput on the same hardware while maintaining full Redis and Memcached API compatibility. Its built-in vector search support enables efficient vector similarity queries at production scale, making it a drop-in replacement for Redis that handles both conventional caching and AI-native workloads without separate infrastructure.
INSTALL
docker run --network=host docker.dragonflydb.io/dragonfly/dragonflyINTEGRATION GUIDE
1. Cache LLM responses and embeddings at high throughput with Redis-compatible APIs and no code changes
2. Run vector similarity searches for RAG pipelines and semantic caching using Dragonfly's FT.SEARCH and TV.SEARCH commands
3. Serve real-time ML features and model inference results with sub-millisecond latency at high concurrency
4. Replace Redis or Memcached deployments in AI stacks to eliminate single-threaded throughput bottlenecks
TAGS
vector-searchredis-compatiblecachingin-memory-databasesimilarity-search