Cloudflare Vectorize vs Vearch

Cloudflare Vectorize is a globally distributed vector database for AI applications built with Cloudflare Workers. It supports semantic search, recommendations, classification, anomaly detection, and supplying context to large language models. Vectors can come from Workers AI models or services such as OpenAI. Workers connect to indexes through a Vectorize binding configured in Wrangler or the Cloudflare dashboard, while a REST API can insert and query vectors. Search results can reference images in R2, documents in KV, and user profiles in D1. Metadata filters support indexed string, number, and boolean properties, with up to 10 metadata indexes per index. Each index can hold up to 20 million vectors, with a maximum of 1,536 dimensions at 32-bit precision. Queries return up to 50 results with values or metadata, or 100 without them. Cloudflare lists a free plan and paid usage from $5/mo. The Workers Free plan includes stated monthly queried and stored dimension allowances. Cloud deployment is available through API and web access.

Vearch is a free, open-source, cloud-native distributed vector database for similarity search in AI applications. It searches embedding vectors and can combine vector queries with scalar filtering for metadata conditions. The project describes replication and elastic scale-out and says it can retrieve vectors among millions of objects in milliseconds. Its Router exposes REST operations for upserting, deleting, searching, and querying data. Client SDKs are listed for Python, Go, Java, and Rust, with integrations for LangChain, LlamaIndex, LangChainGo, and LangChain4j. Deployment options include Kubernetes through Helm, Docker Compose in standalone or cluster mode, Docker, and compilation from source. PartitionServer uses Raft-based replication for document partitions, while Gamma is a Faiss-based vector search engine. Vearch is licensed under Apache License 2.0 and supports Linux and macOS self-hosted deployments. Community support is available through GitHub Issues, an email discussion group, and Slack. The VisualSearch demo requires an image retrieval plugin for object detection and feature extraction.