Weaviate

What it is

Weaviate is an open-source vector search engine and database that uses machine learning to vectorize and store data, enabling natural language queries. It supports custom ML models at production scale.

The only vector database that combines native multi-tenancy, automatic index optimization, and support for custom ML models at production scale.

When it fits — and when it doesn’t

Known weaknesses

Each one links to the public issue, changelog or docs page it came from, where we have one.

1. 01

   No native support for Windows operating systems or binaries.

   A long-standing open issue with 165 reactions highlights the lack of Windows support, forcing users to rely on Linux-based environments or workarounds.

   Source: GitHub issue #3315 ↗
2. 02

   Inability to reindex data smoothly, particularly for updating embeddings or inverted indexes.

   Users report there is no smooth way to reindex data, which is a critical limitation for workflows requiring updates to embeddings or inverted indexes.

   Source: GitHub issue #2359 ↗
3. 03

   No support for cross-tenant queries, requiring multiple separate queries to search across tenants.

   Users cannot search vectors across multiple tenants in a single query, necessitating multiple API calls to aggregate results from different tenants.

   Source: GitHub issue #3500 ↗

What it does well

• Built-in vectorization modules

• GraphQL and REST API

• Multi-tenancy

• Hybrid BM25 and vector search

• Schema-based objects

• Horizontal scaling

• LangChain and LlamaIndex integrations

• Self-hosted and managed cloud

Pricing

Free tier

Yes

Paid from

—

Model

Hybrid

Enterprise plan

No

Pricing not independently confirmed — check the vendor’s page.