Qdrant
Rust vector search engine built for filtering-heavy AI workloads
Overview
Qdrant is an open-source vector database and search engine written in Rust, built for high-performance similarity search with rich metadata filtering. Self-host free or run managed Qdrant Cloud; used in production by Mozilla, Disney+ and Bayer.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 11 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- 0.5 vCPU / 1GB RAM / 4GB disk
- ~250K uncompressed vectors (7-8M with binary quantization)
- Community support only
- Vertical and horizontal scaling
- Backup and disaster recovery
- 99.5% uptime SLA
- 10x5 standard support
- SSO authentication
- Private VPC links
- 99.9% uptime SLA (99.95% multi-AZ)
- 24/7 premium support
- Runs in your own VPC or on-prem
- Full data isolation
- Sales-negotiated contract
Use cases
What you can produce with Qdrant
- Vector similarity search with structured metadata filtering in one query
- Rust core with binary/scalar quantization for smaller indexes
- Self-hosted (Apache 2.0), Qdrant Cloud, Hybrid Cloud or Private Cloud
- Free cloud inference with selected embedding models
- SSO and private VPC links on the Premium tier
- 99.9% uptime SLA (99.95% multi-AZ) on Premium
ASEAN Perspective
Qdrant in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Qdrant's pitch is speed with filters attached — G2 reviewers give it a 9.3 for semantic search accuracy and 9.1 for ease of use in head-to-head comparisons, and metadata filtering (combining vector similarity with structured conditions in one query) is the feature that comes up most in production write-ups. Being Rust rather than Python or Go under the hood shows up in benchmarks and in lower memory footprint at scale, which matters once you're indexing tens of millions of vectors. The free tier (0.5 vCPU, 1GB RAM, 4GB disk) covers real prototyping, not just a toy demo. A $50M Series B in March 2026 (AVP, Bosch Ventures, Spark Capital) on top of a 2024 Series A pushed total funding past $87M, plenty of runway for an infra company. Weak spots reviewers mention: no built-in visualization tooling, and the initial ramp for anyone new to vector search concepts. Solid default for teams that need heavy filtering alongside vector search.
What people say
Qdrant scores well in head-to-head G2 comparisons — 9.3 for semantic search relevance and 9.1 for ease of use against category peers — and reviewers consistently cite speed and scalability as reasons they picked it over alternatives. The combination of vector similarity with structured metadata filtering in a single query gets called out repeatedly as the feature that solves real production problems, particularly for e-commerce search or recommendation work where you need to filter by price, category or availability alongside semantic relevance.
Documentation is rated highly, with reviewers noting they could integrate Qdrant into an existing system without extensive hand-holding. The Rust core is a selling point among engineering-heavy teams: reviewers and third-party benchmarks point to lower memory overhead and more predictable latency under load compared with Python-based alternatives, especially once quantization is turned on to shrink index size.
The most common criticism is the lack of built-in visualization or debugging tools — several reviewers say they had to build their own tooling to inspect what the index was actually doing, a gap competitors with a hosted admin UI don't have as badly. The initial learning curve for teams new to vector search concepts (HNSW indexing, quantization trade-offs, payload filtering syntax) also comes up, though reviewers say the docs eventually get you there.
On pricing, Qdrant Cloud moved to resource-based billing (vCPU, RAM, disk, backup storage) rather than per-query or per-vector charges, so costs stay flat regardless of query volume — reviewers who ran high-QPS workloads specifically call this out as cheaper than Pinecone at their scale. The free tier handles roughly 250K uncompressed vectors, or up to 7-8M with binary quantization turned on, generous enough for most prototypes and small production workloads.
Qdrant closed a $50M Series B in March 2026 led by AVP with Bosch Ventures, Unusual Ventures, Spark Capital and 42CAP participating, following a $28M Series A in 2024 — total disclosed funding around $87.8M. On GitHub, qdrant/qdrant carries 33.2k stars under Apache 2.0, ahead of Weaviate and near Chroma in community size, and the company ships new releases on a regular cadence.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Qdrant's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Qdrant unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Qdrant directly →
Spotted something out of date? Suggest an update →
Qdrant in the news
More in Code & Dev Tools
