Qdrant is a vector search engine with filtering, recommendations, and multitenancy.
- Stars
- 35,619
- Updated
- License
- Apache-2.0
- Hosting
- Both
- Category
- Search
Qdrant has the most GitHub stars in this list (36k, Apache-2.0). Qdrant, ParadeDB, and Vespa can be self-hosted. Qdrant lists a hosted option.
Compare the setQdrant is a vector search engine with filtering, recommendations, and multitenancy.
ParadeDB adds full-text search, BM25 ranking, and vector retrieval to PostgreSQL.
Often instead of Elasticsearch
WebsiteA search and serving platform for indexing data, ranking results, and vector search.
The catalog lists 3 reviewed open-source alternatives to Pinecone for storing and querying vector embeddings: Qdrant (Apache-2.0, hosted or self-hosted), ParadeDB (AGPL-3.0, self-hosted), and Vespa (Apache-2.0, self-hosted). GitHub metadata last checked November 18, 2026.
Start with your workflow for storing and querying vector embeddings. The alternatives below can cover different parts of it; a listing here does not establish complete feature parity with Pinecone.
| Attribute | Qdrant | ParadeDB | Vespa |
|---|---|---|---|
| Purpose | Qdrant is a vector search engine with filtering, recommendations, and multitenancy. | ParadeDB adds full-text search, BM25 ranking, and vector retrieval to PostgreSQL. | A search and serving platform for indexing data, ranking results, and vector search. |
| Documented capabilities | Faceting, Recommendation, Discovery, Search Relevance Tuning, Multitenancy, Observability | Full-text search, BM25 scoring, Vector search | Data indexing, Ranking models, Vector search |
| Best-fit workflow | Qdrant is a vector search engine with filtering, recommendations, and multitenancy. | ParadeDB adds full-text search, BM25 ranking, and vector retrieval to PostgreSQL. | A search and serving platform for indexing data, ranking results, and vector search. |
| Check before choosing | Check the project's current installation requirements and license terms before adoption. | Check the project's current installation requirements and license terms before adoption. | Check current installation requirements and edition limits in the linked repository. |
| Deployment options | Docker | See repository installation guide | See repository installation guide |
| GitHub stars | 35,619 | 9,539 | 7,185 |
| License | Apache-2.0 | AGPL-3.0 | Apache-2.0 |
| Self-hosted | Yes | Yes | Yes |
| Docker noted | Yes | Not recorded | Not recorded |
| Cloud / hosted | Yes | Not recorded | Not recorded |
| Primary language | Rust | Rust | Java |
| Last commit | 2026-11-17 | 2026-11-18 | 2026-11-17 |
| Difficulty | Not assessed | Not assessed | Not assessed |
| Content source | Editorial catalog entry | Editorial catalog entry | Editorial catalog entry |
| Metadata checked | 2026-11-18 | 2026-11-18 | 2026-11-18 |
The table shows up to six catalog entries. Missing information means not recorded, not unsupported. Capability lists are not hands-on benchmarks or guarantees of Pinecone compatibility.
A vector search engine with filtering and multitenancy.
Check first: Check the project's current installation requirements and license terms before adoption.
Qdrant documentation and sourceApache-2.0 · Both · Search
Vector retrieval and full-text search inside PostgreSQL.
Check first: Check the project's current installation requirements and license terms before adoption.
ParadeDB documentation and sourceAGPL-3.0 · Self-hosted · Search
Indexing, ranking, and vector search in one serving platform.
Check first: Check current installation requirements and edition limits in the linked repository.
Vespa documentation and sourceApache-2.0 · Self-hosted · Search
Use this proposed pilot to test storing and querying vector embeddings. It is an evaluation plan, not a completed product test.
Repository activity is a signal, not a security or quality guarantee.