#1515th of 22 in Vector databases
infinity
Database for hybrid dense, sparse, tensor and full-text search
- Stars
- 4.8k
- License
- Apache-2.0
- Last commit
- Nov 2026
- Last release
- Nov 2026
- Language
- C++
4,451 on 22 Apr 20264,764 on 11 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Infinity is a C++ database that runs hybrid search over dense vectors, sparse vectors, tensors and full-text, with filtering on structured columns. It supports RRF, weighted-sum and ColBERT rerankers, and is used from a Python SDK (embedded or over the network) or an HTTP API. The README reports 0.1 ms latency and 15K+ QPS on million-scale vector sets, and 1 ms latency on full-text search over 33M documents.
Who it is for: Developers building RAG or search apps who want a self-hosted hybrid database
Strengths
- One query combines dense, sparse, tensor and full-text search with filters
- Ships RRF, weighted-sum and ColBERT rerankers
- Single binary with no external dependencies; also embeddable as a Python module
- Python SDK and HTTP API available
Weaknesses
- Requires an x86_64 CPU with AVX2; no ARM support listed
- Quickstart uses the nightly Docker tag rather than a pinned release
- Benchmark figures are self-reported in the README
- Windows use requires WSL/WSL2 with host networking
What it needs
- no GPU
- Models: dense embeddings, sparse embeddings, tensors, ColBERT reranking
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in Vector databases
See all 22| Rank | Project | Score |
|---|---|---|
| 1 | MilvusDistributed vector database with dense, sparse and hybrid search at scale | 90 out of 100 |
| 2 | QdrantRust vector database with payload filtering, REST and gRPC | 85 out of 100 |
| 3 | MeilisearchRust search engine with typo tolerance, filtering and hybrid semantic search | 84 out of 100 |
| #4 | faissC++ library for similarity search and clustering of dense vectors | 81 out of 100 |
| #5 | leannVector index that recomputes embeddings on demand to cut storage | 79 out of 100 |