Open-Source Databases · slothflowlabs

duckle

Duckle is an open-source desktop ETL/ELT studio for building and running visual data pipelines locally using DuckDB. It combines 290+ data connectors, a visual canvas, dbt integration, and an on-device AI assistant into a single ~65 MB application with no cloud dependency.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/slothflowlabs/duckle

757

GitHub stars

Forks

Rust

Primary language

Apache-2.0

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	slothflowlabs/duckle
Owner	slothflowlabs
Primary language	Rust
License	Apache-2.0 — OSI-approved
Stars	757
Forks	50
Open issues	43
Latest release	v0.5.4 (2026-07-05)
Last updated	2026-07-08
Source	https://github.com/slothflowlabs/duckle

What duckle is

Built in Rust with Tauri 2 and React, Duckle executes pipelines by compiling visual node graphs to SQL and running them through DuckDB. It includes 74+ sources, 126+ transforms, 58+ sinks, local LLM integration (Qwen 2.5 Coder 1.5B via llama.cpp), dbt orchestration, and a scheduler for production deployment.

Quickstart

Get the duckle source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/slothflowlabs/duckle.git2cd duckle3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Local-first data transformation for small-to-medium teams

Organizations needing to build, test, and iterate on ETL/ELT pipelines without cloud infrastructure lock-in or ongoing SaaS costs. Audit-friendly, single-file binary deployable on air-gapped machines.

Rapid pipeline prototyping with AI assistance

Data engineers can describe pipeline requirements in natural language to Duckie (local LLM); the assistant generates valid pipeline JSON for immediate canvas insertion, reducing manual authoring overhead.

Multi-source data integration at scale

Projects requiring connectivity to 290+ sources (legacy databases, SaaS APIs, data warehouses, vector stores, streaming brokers) in one tool, with visual monitoring and generated SQL transparency.

Implementation considerations

Beta status: latest release v0.5.4 is from 2026-07-05 with active development; expect breaking changes or gaps in edge cases. Review open issues (43 tracked) and GitHub discussions before production adoption.
Offline/air-gapped deployment: requires pre-download of ~1.1 GB LLM model and DuckDB engines; verify distribution mechanism and storage in restricted networks.
Workspace format: pipelines stored as plain files in user-selected folder; version control (Git integration mentioned) is user responsibility; backup and recovery strategy essential.
Connector maturity: 290+ connectors span many platforms; validate critical connectors in test environment and review release notes for known gaps (e.g., SAP support only added in v0.5.4).
Performance tuning: DuckDB execution is fast for OLAP workloads, but pipeline logic complexity, data volume, and join strategy will impact runtime; documentation mentions performance tips but not provided here.

When to avoid it — and what to weigh

Need production support and SLAs — Duckle is marked 'beta' with 757 stars and ~14 weeks of visible history. No indication of commercial support tier, warranty, or enterprise SLA. Suitable for internal or exploratory use only.
Require extensive connector ecosystem for proprietary or legacy systems — While 290+ connectors are claimed, not all may be production-ready or fully tested. Critical integrations should be validated against release notes and GitHub issues before adoption.
Strict compliance/governance requirements — Local-first execution and file-based workspaces lack built-in audit logging, access control, or compliance framework (HIPAA, SOC 2, etc.). Not designed for regulated environments without custom hardening.
Real-time or sub-second latency pipelines — Duckle targets batch and near-real-time workflows via DuckDB. Streaming ingestion or microsecond latency requirements should evaluate dedicated streaming platforms.

License & commercial use

Dual-licensed under MIT OR Apache-2.0 (Apache License 2.0). Both are permissive OSI-approved licenses. Apache-2.0 explicitly permits commercial use, modification, and distribution with proper attribution and license inclusion. Inspection of source code for third-party dependencies and their licenses is recommended.

Apache-2.0 is a permissive license that permits commercial use, provided the license text is included in distributions. No vendor lock-in, no proprietary restrictions stated. However, no commercial support, indemnification, or warranty is mentioned. Internal use and bundling are permitted; distribution to customers should include Apache notice. Verify third-party dependency licenses (e.g., DuckDB, Tauri, llama.cpp) for compliance in your use case.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Adequate
License clarity	Clear
Deployment complexity	Low
DEV.co fit	Good
Assessment confidence	Medium

Security considerations

No security audit, CVE history, or threat model provided. Local-first architecture mitigates cloud-side data exposure; execution runs on user's CPU/storage. Risks to evaluate: (1) Tauri/Electron/browser engine vulnerabilities in desktop app, (2) DuckDB and third-party dependency security patches, (3) LLM model (Qwen 2.5) trustworthiness and updates, (4) plaintext workspace files and credential storage (API keys, DB passwords in pipeline definitions), (5) no built-in RBAC or audit logging. Use only in trusted network environments; rotate credentials regularly; validate third-party connectors before production.

Alternatives to consider

Apache Airflow

Mature, production-grade Python workflow orchestrator with extensive plugin ecosystem. Requires DevOps infrastructure (scheduler, metadata DB) and stronger Python/DevOps skills; steeper learning curve but stronger enterprise governance.

dbt Cloud / dbt CLI + Git

Focused on dbt-native transformations with integrated CI/CD and job scheduling. Narrower scope than Duckle (no multi-source connectors, visual UI); lower TCO for SQL-only teams; cloud or self-hosted options available.

Stitch / Talend Cloud

Commercial cloud ETL/ELT platforms with broad connector support, managed infrastructure, and SLAs. Higher cost and cloud lock-in; suitable for teams requiring vendor support and compliance frameworks.

Software development agency

Build on duckle with DEV.co software developers

Download the beta, test with a sample pipeline, and assess fit for your use case. Join the Discord community for support. Review the roadmap and open issues to understand current gaps.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

dbx duckdb-wasm tad

Related on DEV.co

Explore the category and the services that help you build with it.

Open-Source Databases AI development API development Enterprise software Talk to DEV.co

duckle FAQ

Can I use Duckle in production?

Yes, with caveats. Duckle is marked 'beta' (v0.5.4); internal/non-critical use is safe. For customer-facing or mission-critical workflows, evaluate: (1) your tolerance for breaking changes, (2) lack of commercial SLA, (3) absence of formal audit/compliance framework, (4) team's ability to maintain local deployment and debug issues independently.

Do I need an API key or internet connection?

No. Duckie AI (Qwen LLM) runs locally via llama.cpp after one-time download (~1.1 GB). Pipelines execute entirely on your CPU using local DuckDB. Internet only needed for data source/sink connectors (e.g., pulling data from cloud APIs or warehouses).

How are workspaces stored and versioned?

Workspaces are plain files in a folder you choose. In-app Git integration (GitHub/GitLab) is available for version control. File-based storage allows diff/branch/audit but requires you to manage backup, recovery, and CI/CD integration manually.

What happens if I hit an unsupported connector or edge case?

Duckle has 290+ connectors, custom code nodes, and custom AI transform options; you can extend via MCP server integration or contribute to the open-source project. Community support via Discord; no vendor SLA. Early-stage projects may have gaps; validate critical connectors before deployment.

Work with a software development agency

DEV.co helps companies turn open-source tools like duckle into production software. Our software development services cover the full lifecycle — architecture, web development, integration, and maintenance — delivered by software developers and web developers who ship. Engage our software development agency to implement or customize it for your open-source databases stack.

Evaluate Duckle for Your Data Pipeline Needs

Download the beta, test with a sample pipeline, and assess fit for your use case. Join the Discord community for support. Review the roadmap and open issues to understand current gaps.

Talk to DEV.co Schedule a Consultation