Weaviate
Separate Weaviate database capacity from embeddings and Query Agent costs when budgeting retrieval.
- Starting price
- Flex from USD 45/month
- Free access
- One cluster; 100,000 objects
- Billing unit
- Database resources plus AI-service usage
- Platform
- Managed cloud or open-source deployment
- Verified
- 30 Oct 2026
Overview
What you provide
- Provide a schema, authorised records or vectors, and the filters and queries your retrieval task requires.
What you get
- The database returns matching records for vector, keyword or hybrid search; configured AI services add their own outputs.
Choosing this tool
Weaviate is a vector database for semantic and hybrid retrieval in AI applications. The managed Free cluster is a practical small-data starting point, but its one-collection limit differs materially from a production multi-collection design. Compare Milvus for a self-operated vector database or Qdrant for another vector-search stack. Decide using a representative retrieval evaluation, not a claim that adding a database automatically improves answer quality.
A decision worked through
Combining the USD 45 Flex minimum with the listed USD 30 Query Agent monthly plan gives a USD 75/month starting subtotal before embeddings, extra requests, taxes or resource scaling. Using only the database does not require assuming that extra Query Agent subscription.
Weaviate alternatives for managed vector and hybrid retrieval for ai
Compare Weaviate with Milvus, Qdrant for this workflow, including the billing unit and the work left to your team.
| Product | Price / unit | Free allowance | Input → output | Key limit |
|---|---|---|---|---|
| WeaviateRead profileOfficial source ↗ | Flex from USD 45/month | 100,000 objects on managed Free | Objects and queries become retrieval matches. | AI-service usage is separately metered. |
| MilvusRead profileOfficial source ↗ | Apache-2.0; no licence fee | Self-hosted software without cloud credits | Vectors and filters return ranked neighbours. | Deployment operations are separate from the Lite demo. |
| QdrantRead profileOfficial source ↗ | Apache-2.0 software; infrastructure extra | No software licence charge for self-hosting | Vectors and filters yield matching records. | You operate the self-hosted database and embedding pipeline. |
Setup & workflow
- Create a test collection and load a representative, permitted corpus.
- Compare vector and hybrid results with expected evidence for known questions.
- Estimate resources and AI-service usage separately before enabling production traffic.
Requirements & installation
- Choose a deployment mode and a sample corpus; embedding dimensions, object size and expected concurrency affect resources.
Limits & review
- Free lacks backup retention and is limited to one collection. Capacity, embedding requests, Query Agent calls and answer generation are separate operating concerns.
Your part
- Evaluate retrieved evidence against labelled questions before connecting it to an answer-generating model.
Sources & verification5
This profile is based on official sources, not a hands-on product test.
Vendor descriptions and demos document advertised features. Editorial guidance is based on these sources.
- Official Weaviate pricingweaviate.io
- Official Weaviate productweaviate.io
- Weaviate official referenceweaviate.io
- Milvus official referencemilvus.io
- Qdrant official referencegithub.com
