SQL-LIKE VECTOR SEARCH QQL — SQL for Qdrant
A single-binary CLI, SDK, and gateway for operating Qdrant vector databases with a SQL-like query language. Write search logic like SQL. Ship it like a binary. Run it anywhere.
# Install QQLcurl -fsSL https://raw.githubusercontent.com/srimon12/qql-go/main/install.sh | sh
# Connect to Qdrantqql-go connect --url http://localhost:6334
# Create collection and load dataqql-go exec "CREATE COLLECTION docs HYBRID"qql-go exec "INSERT INTO docs VALUES {'id': 1, 'text': 'Qdrant stores vectors'} USING HYBRID"
# Hybrid query searchqql-go exec "QUERY 'vector database' FROM docs LIMIT 5 USING HYBRID"Why QQL?
A query language and tooling layer built specifically to make vector database operations simple, governed, and highly performant.
Unified Querying
Dense, sparse, hybrid, and multivector retrieval modes are all unified under a single SQL-like QUERY syntax. No need to write complex nested JSON or gRPC structures.
Single Binary Portability
Compiled in Go with zero runtime dependencies. A single executable that runs natively on Linux, macOS, and Windows. No SDK updates or environment versioning drift.
In-Database Algebra
Perform expression algebra, boosting, and filtering directly within the Qdrant query pipeline. Eliminates the need for expensive client-side post-processing and sorting.
Security & Governance Gateway
Secure production databases with built-in JWT authentication, rate limiting, audit logging, and dynamic tenant filter injection enforced at the AST parser level.
Integrate in Any Stack
Official client SDKs are currently in development. Integrate QQL into your workflow today using our CLI and gateway APIs.
Client SDKs
Type-safe client SDKs are available in the source repository but not yet published to package registries:
TypeScript (In Source) Go SDK (In Source) Python Client (In Source) Active Integrations
Deploy the QQL Gateway and interact with Qdrant today using our stable communication protocols:
REST API gRPC / Connect RPC Single-Binary CLI Frequently Asked Questions
Answers to common operational, architectural, and security questions about QQL.
What is QQL?
QQL (Qdrant Query Language) is a SQL-like interface, single-binary CLI, and security gateway designed for operating Qdrant vector databases. It allows developers to express dense, sparse, and hybrid retrieval logic using a declarative SQL dialect.
Does the QQL Gateway support multi-tenancy?
Yes. The QQL Gateway intercepts queries, validates JWT claims, and structurally injects tenant isolation filters at the AST (Abstract Syntax Tree) compiler level. This prevents cross-tenant data leaks and prompt injection attacks in AI agent workstreams.
Is QQL open source?
Yes, QQL is fully open source under the MIT license. The gateway, SDKs, and CLI are free to run, self-host, and modify.
Does QQL require re-indexing Qdrant?
No. QQL compiles SQL-like queries dynamically to Qdrant's native API calls (gRPC/REST). Your existing collections, points, and payload indexes in Qdrant remain untouched.