# Vector databases > RAG, semantic search and agent memory all come down to one query: find the nearest vectors, fast, with filters. Pinecone is managed only. Qdrant, Weaviate and Chroma are open source with a cloud on top. Glider keeps every write in your own S3 bucket. KiviDB adds vector search to a Redis-compatible store. 6 tools, in the order they appear in the Digidle Tools directory. This is not a ranking. - Page: https://digidle.com/best/vector-databases/ - All best-of lists: https://digidle.com/best/ ## Tools - [Glider](https://digidle.com/tools/glider/): Open-source single-node vector database in Rust that makes every acknowledged write durable in S3-compatible storage, with filtered nearest-neighbor search over an HTTP API, a Python client and an MCP memory server for agents. Official site: https://gliderdb.com. Alternatives: https://digidle.com/alternatives/glider/. - [KiviDB](https://digidle.com/tools/kividb/): Open-source (AGPL-3.0), multi-threaded Redis and Valkey alternative written in Rust that speaks RESP2/RESP3, so existing Redis clients work unchanged, with full-text and vector search built in; self-host it or use KiviDB Cloud. Official site: https://kividb.io. Alternatives: https://digidle.com/alternatives/kividb/. - [Pinecone](https://digidle.com/tools/pinecone/): Fully managed vector database: serverless dense, sparse and full-text indexes on AWS, GCP or Azure, with hosted embedding and reranking models, Pinecone Assistant for chat and agent apps, and bring-your-own-cloud on Enterprise. Free Starter plan. Official site: https://www.pinecone.io. Alternatives: https://digidle.com/alternatives/pinecone/. - [Qdrant](https://digidle.com/tools/qdrant/): Open-source (Apache-2.0) vector database and search engine written in Rust: filtered vector search you can self-host, or run on Qdrant Cloud on AWS, GCP or Azure with a free 1 GB cluster, usage-based dedicated clusters, and hybrid or private cloud on your own infrastructure. Official site: https://qdrant.tech. Alternatives: https://digidle.com/alternatives/qdrant/. - [Weaviate](https://digidle.com/tools/weaviate/): Open-source (BSD-3-Clause) vector database that stores objects and vectors together, with hybrid search, filtering, multi-tenancy and vector compression. Self-host it or use Weaviate Cloud, which adds hosted embeddings and a Query Agent; free cluster for 100,000 objects. Official site: https://weaviate.io. Alternatives: https://digidle.com/alternatives/weaviate/. - [Chroma](https://digidle.com/tools/chroma/): Open-source (Apache-2.0) search infrastructure for AI: vector, full-text and metadata search. Self-host it or use serverless Chroma Cloud, billed per GiB written, stored and queried, with free credits to start. Official site: https://www.trychroma.com. Alternatives: https://digidle.com/alternatives/chroma/. ## At a glance | Tool | Open source | Where it runs | Free option | Price | | --- | --- | --- | --- | --- | | [Glider](https://digidle.com/tools/glider/) | Yes, Apache-2.0. | Self-hosted, one server in Docker, with S3-compatible storage as the source of truth. No replication. | Free to self-host. | No paid plan listed. You pay for your own server and bucket. | | [KiviDB](https://digidle.com/tools/kividb/) | Yes, AGPL-3.0. | Self-hosted, or KiviDB Cloud on AWS, Google Cloud, Azure or your own cloud account. | Free to self-host. | Cloud: Essentials, Pro and Scale tiers, billed by the hour. See pricing. | | [Pinecone](https://digidle.com/tools/pinecone/) | No. | Managed on AWS, GCP or Azure. Bring-your-own-cloud runs it in your account. | Starter: 2 GB storage, 2M write units and 1M read units a month, up to 5 indexes, us-east-1 only. | Builder $20/mo flat. Standard $50/mo minimum, then pay as you go. Enterprise $500/mo minimum. | | [Qdrant](https://digidle.com/tools/qdrant/) | Yes, Apache-2.0. | Self-hosted, Qdrant Cloud on AWS, GCP or Azure, or Hybrid and Private Cloud on your own infrastructure. | Free tier: one-node cluster with 0.5 vCPU, 1 GB RAM and 4 GB disk. | Standard is usage-based, billed hourly on vCPU, memory, storage and backups. Premium has a minimum spend. | | [Weaviate](https://digidle.com/tools/weaviate/) | Yes, BSD-3-Clause. Some enterprise features need a license key. | Self-hosted, or Weaviate Cloud: shared on AWS or GCP, dedicated on AWS, GCP or Azure. | Free: 1 cluster, 100,000 objects, 1 GB memory, 10 GB disk. | Flex from $45/mo, pay as you go. Premium is a prepaid contract from $400/mo. | | [Chroma](https://digidle.com/tools/chroma/) | Yes, Apache-2.0. | Self-hosted, or serverless Chroma Cloud. BYOC and single-tenant clusters on Enterprise. | Starter: $0/mo plus usage, with $5 in free credits. | $2.50 per GiB written, $0.33 per GiB stored a month, $0.0075 per TiB queried. Team $250/mo with $100 credits. Enterprise custom. | Checked 2026-10-30 on vendor pages: [Pinecone pricing](https://www.pinecone.io/pricing/), [Qdrant pricing](https://qdrant.tech/pricing/), [Qdrant on GitHub](https://github.com/qdrant/qdrant), [Weaviate pricing](https://weaviate.io/pricing), [Weaviate license](https://github.com/weaviate/weaviate/blob/main/LICENSE), [Chroma pricing](https://www.trychroma.com/pricing), [Chroma on GitHub](https://github.com/chroma-core/chroma), [Glider](https://gliderdb.com), [Glider on GitHub](https://github.com/omerfeyzioglu/glider), [KiviDB pricing](https://kividb.io/pricing). ## Pick it if - [Pinecone](https://digidle.com/tools/pinecone/): You want nothing to run, with embeddings, reranking and search from one vendor. - [Qdrant](https://digidle.com/tools/qdrant/): You want open source you can self-host today and the same engine managed later, even on your own infrastructure. - [Weaviate](https://digidle.com/tools/weaviate/): You want hybrid search and multi-tenancy, with embeddings hosted next to the data. - [Chroma](https://digidle.com/tools/chroma/): You want to start on your laptop and move to a serverless cloud billed per GiB. - [Glider](https://digidle.com/tools/glider/): You want your vectors in your own S3 bucket, safe through a crash, and one server is enough. - [KiviDB](https://digidle.com/tools/kividb/): You already run Redis and want vector search without adding another database. --- Source: Digidle Tools (https://digidle.com/). Full index: https://digidle.com/llms.txt