AI Potluck
Infrastructure / Storage & Retrieval

Weaviate

Weaviate

Weaviate is an open-source vector database that stores objects and vectors together, combining vector search with structured filtering, keyword and hybrid search behind a GraphQL and REST API. It runs embedded for development, self-hosted as a Docker or Kubernetes cluster, or on the vendor's hosted service, with optional modules that call out to embedding and reranking models.

No PyPI artifact is declared on purpose. weaviate-client (80M downloads a month, published from weaviate/weaviate-python-client) is the Python client for a self-hosted or hosted Weaviate instance, not the server; installing it gives you a way to reach a Weaviate, not a Weaviate. Declaring it would make the client's count the product's adoption on the next warehouse refresh, so the package is recorded in the resolution ledger instead and adoption reads from stars. Verified 2026-09-02 via the GitHub API and the LICENSE body.

Openness

5 high confidence
5.0
license
BSD-3-Clause(OSI)
source
public
core-gated
ungated

BSD-3-Clause license body confirmed. The repository is public and unarchived and builds the whole server, and the README describes no paid tier, enterprise edition or license-gated build beside it - Weaviate Cloud is a hosted service sold beside the BSD-licensed core rather than a piece withheld from it. So source is public and the core ungated.

Adoption

3 low confidence
3.0

16,776 GitHub stars, which lands in the >10K stars band of the stars scale, level 3. No package artifact is declared - the PyPI weaviate-client package is a client for a running Weaviate instance rather than the server, so it is left undeclared rather than declared and ignored (see the product comments). Weaviate is deployed as a Docker or Kubernetes cluster with no published deployment count, which leaves the stars scale and its cap of 3.

Capability

4 medium confidence
4.0

Banded on the category feature matrix as complete retrieval system, distributed or cluster-capable. Placed one band below the vespa anchor, on a matrix that bands on how much of the persist-index-retrieve path a product provides itself, at what scale and over how many retrieval modes. Weaviate's vectorizer and reranker modules are optional integrations that call an external or containerized model rather than the core serving engine hosting and evaluating one itself, which is the line the top rung holds - the same distinction that keeps Milvus's and Qdrant's rerank functions at this band rather than the one above.

  • https://github.com/weaviate/weaviate/blob/main/README.md recorded 2026-09-02

    README documents a cloud-native vector database with replication, sharding and multi-tenancy, and an optional modules system (text2vec, reranker, generative) that integrates external or containerized models rather than hosting inference inside the core query engine.

Verified 2026-09-02