Quickwit

Quickwit is an open-source, cloud-native search and analytics engine built in Rust, designed specifically to store and query data on object storage such as Amazon S3, MinIO, and Ceph with sub-second response times. The project has joined Datadog as of 2025 while remaining open source. Its custom file format minimizes I/O requests, and intelligent scheduling maximizes throughput on cloud storage backends, achieving up to 90-percent cost reduction compared to traditional Elasticsearch deployments. The engine is powered by Tantivy, the fastest full-text search library in the Rust ecosystem, and includes vectorized processing with SIMD instructions. Quickwit provides native ingestion from Kafka and Kinesis streams, schemaless indexing, native OpenTelemetry and Jaeger trace support, GDPR-compliant targeted deletions, and retention policies for log lifecycle management. Customers include Binance for high-volume log search, PostHog for product analytics, Fly.io for infrastructure metrics, Contentsquare for session replay search, Stedi for EDI document search where Founder and CEO Zack Kanter reports 80-percent cost savings versus Elasticsearch, Rho for observability with 90-percent tracing cost reduction, Matter Labs for ZK rollup indexing, and World Wide Technology for enterprise data lakes. The Tantivy library underlying the project is independently used by Milvus and Zilliz for vector search, Databend for cloud data warehousing, ParadeDB for PostgreSQL full-text search, LanceDB and MyScale for AI similarity search, and TabbyML for code search. Deployment options include single-node, multi-node clusters, and fully managed cloud.