Weaviate

Weaviate is a vector database for developers building search, RAG, and agent applications. It suits teams that need control over retrieval and deployment. Its search and hosting options are substantial, but operating a database—and moving beyond a tightly capped free tier—takes technical commitment.

Overview

Weaviate stores, indexes, and searches high-dimensional vectors as infrastructure for applications, rather than as a finished tool for end users. Teams can use its hosted service or run it themselves locally or with a cloud provider, including through Docker or Kubernetes. That flexibility helps developers choose how much to manage, but Weaviate is best for teams prepared to work with database infrastructure.

Key features

Search and retrieval

Hybrid search combines vector search with BM25 keyword matching; semantic and keyword search are also supported. Metadata filtering and sparse vectors add retrieval options, while HNSW, Flat, Dynamic, and HFresh are available index types. This range suits applications that need to balance concept-based matches with exact terms, though it brings choices that developers must make for their use case.

RAG, controls, and integrations

Weaviate describes out-of-the-box RAG for interacting with proprietary data through machine-learning models. Vectorizer modules, compression, native multi-tenancy, tenant isolation, and configurable backups extend the database’s data and retrieval controls. Integrations include AWS, Google Cloud, Azure, OpenAI, Anthropic, Cohere, Hugging Face, and Ollama, giving teams a broad set of cloud and model options.

Weaviate Cloud encrypts data in transit and at rest, supports role-based access controls, and provides automated daily backups. Weaviate Cloud is SOC 2 audited; Enterprise Cloud on AWS is HIPAA compliant. These provisions matter to teams evaluating hosted deployment, but they do not remove the need to assess a deployment against its own security and compliance requirements.

Deployment and platforms

Weaviate supports API, Linux, self-hosted, and web use. Windows use is possible through containers such as Docker or WSL, but native Windows support is not offered. Its SDK languages include Python, TypeScript/JavaScript, Go, Java, and C#, a useful spread for development teams working in those ecosystems.

Pricing

Free

The Free plan costs nothing, billed Forever. It includes one cluster per user, 100,000 objects, 1 GB memory, 10 GB disk, one collection, and up to three tenants. Embeddings are capped at 2,000 requests per day and Query Agent at 1,000 requests per month. Basic Support is included, but there are no backups or high-availability replication. That makes Free suited to learning and limited evaluation, not workloads that need resilience or more room to grow.

Flex

Flex starts at $45 a month, billed Pay-as-you-go, monthly, with no commitment. It provides unlimited objects, up to 1,000 collections, a shared cluster, and 99.5% uptime, with Standard Support and next-business-day Sev 1 response. The move to a shared cluster and stated uptime make this a practical step beyond Free for expanding workloads, while teams needing stronger availability or a faster Sev 1 response should consider Premium.

Premium

Premium costs $400 a month, billed Prepaid contract. It includes unlimited objects, 1,000 collections, 45 days of backup retention, and up to 99.95% uptime, plus enterprise support with Sev 1 response as fast as one hour. It fits teams that value retained backups, higher stated uptime, and faster support; the prepaid contract is a more committed choice than Flex’s monthly, no-commitment billing.

Platforms

Weaviate’s API, Linux, self-hosted, and web options accommodate teams building applications around a database they can host or manage themselves. Container-based Windows use is possible, but those who require native Windows support should look elsewhere.

Who it's for

Weaviate is aimed at developers, and its deployments serve startups, scale-ups, and enterprises. It is a strong fit for teams building search, RAG, or agent systems that need hybrid retrieval, multiple deployment choices, or integrations across cloud and model providers. It is less suitable for people looking for a finished search application or teams without the capacity to configure and operate database infrastructure.

Pros and cons

• Pros: Hybrid, semantic, and keyword search give developers several ways to retrieve relevant data.
• Pros: Hosted and self-managed deployment, plus Docker and Kubernetes options, let teams choose where to run it.
• Pros: Flex has unlimited objects, 1,000 collections, and monthly no-commitment billing starting at $45 a month.
• Cons: Free is limited to 100,000 objects, one collection, and three tenants, with no backups or high-availability replication.
• Cons: Native Windows support is absent; Windows users must use containers such as Docker or WSL.
• Cons: Premium’s $400 a month prepaid contract is a larger commitment than Flex.

Alternatives

Consider Chroma if Apache 2.0 licensing and broad platform coverage, including mobile and desktop systems, are priorities. Choose Qdrant if its free cloud tier’s single-node setup and specified CPU, memory, and disk allocation fit your needs. Vespa is an option for readers willing to consider paid plans and a free trial. Zvec is a free, Apache 2.0-licensed in-process vector database for those who prefer that model. EmergentDB offers a free Community plan with unlimited vectors, queries, and writes. ZeusDB Vector Database is a free self-hosted option with a Python API, HNSW search, persistence, and metadata filters. Cloudflare Vectorize may suit readers whose needs fit its Workers Free quotas for stored and queried vector dimensions. Epsilla has a free tier with one team member, one project, and 50 messages per month, with data cleared after three months.

Browse Vector Databases, Search Databases, or Database Software to compare more options.

Verdict

Choose Weaviate if you are a developer building search, RAG, or agent applications and want hybrid retrieval with a choice of hosted or self-managed infrastructure. Its strongest reason to choose it is the combination of configurable search and deployment flexibility; its main reason to look elsewhere is the technical overhead, compounded by a free tier without backups or high-availability replication.