dbt does one job: it turns raw tables already sitting in a warehouse into tested, documented models written as SQL. Teams look elsewhere when they want a single product that also loads the data, when per-seat pricing stops fitting a growing analytics team, when the vendor's merger changes their risk picture, or when the people building pipelines do not write SQL.
Sebastiaan SmitsFounderShort answer: The first three alternatives to dbt in our ETL and data integration ranking are Weld, Keboola and KNIME. Smaller vendors fill the top three places where the category offers them, and the household names start at place 4. Choose Weld if you want loading and SQL models in one Danish product at a readable price. Each of the 11 alternatives sits in Best ETL Tools in 2026 and is described against dbt.
Vendors can pay for more visibility on this website, for example a higher place in a ranking or more room on a page. Commercial relationships can also affect which products we feature and how prominently. Every vendor link is marked the same, whatever the arrangement behind it. How that works.
dbt's own introduction says it works alongside ingestion tools and transforms data inside your cloud data platform. Getting data in is somebody else's job, so a team that picks dbt still has to buy or build the extraction layer, and then wire two products together.
The free Developer plan has one seat and 3,000 successful models a month, and its FAQ excludes API access. Starter costs $100 per user per month for five developer seats and 15,000 models. Single sign-on, role-based access control and SCIM appear only on Enterprise, which is quoted, and Starter is billed monthly by credit card.
dbt Labs announced its merger with Fivetran on 13 October 2025. The announcement says dbt Core and Fusion keep their licences and that both products continue, so nothing forces a move. Buyers who chose dbt partly for vendor independence still tend to reassess a supplier after a merger.
| Alternative | Headquarters | Free option | Pricing model | Switch if | Our rank |
|---|---|---|---|---|---|
| Weld | Denmark | No free tier | Published | you want loading and SQL models in one Danish product at a readable price | #3 |
| Keboola | Czechia | Yes | Published | a small team without a data engineer needs the whole flow in one place | #1 |
| KNIME | Switzerland | Yes | Published | analysts who do not write SQL should build the transformations visually | #14 |
| Easy Data Transform | United Kingdom | No free tier | See profile | the work is file wrangling on one machine and a subscription feels excessive | #19 |
| Dataiku | France / United States | No free tier | Published | transformation is one step in a wider machine learning workflow | #15 |
| Matillion | United Kingdom | No free tier | On request | you want to keep warehouse-side processing but hand the design to a visual editor | #6 |
| Microsoft Power Query and Dataflow Gen2 | United States | No free tier | See profile | your reporting already lives in Power BI or Excel and you want no extra licence | #17 |
| Azure Data Factory | United States | No free tier | Published | you sit on Azure and also need to reach on-premise databases | #10 |
| AWS Glue | United States | No free tier | Published | your data platform is AWS and somebody on the team knows Spark | #11 |
| Qlik Talend Cloud | United States | No free tier | On request | auditors want data quality evidence attached to every pipeline | #9 |
| Informatica IDMC (Salesforce) | United States | Yes | Published | a large regulated estate with legacy systems has to clear procurement | #12 |
Denmark · Published
Weld comes from Copenhagen and bundles extraction, SQL transformation and lineage in one product, which removes the second purchase that dbt implies. Tiers are published from $99 a month by monthly active rows and connectors, with a 14-day trial. Complicated model graphs strain its transformation layer, and its connector list is shorter than the older vendors'.
Switch from dbt if you want loading and SQL models in one Danish product at a readable price.
Czechia · Published
Keboola, from Czechia, runs extraction, transformation and orchestration on one platform in an AWS, Azure or GCP region you choose. Billing counts processing minutes: the free plan includes 60 a month and top-ups are $0.14 a minute. The enterprise plan adds a data catalogue. Minute-based costs are hard to estimate before real loads run.
Switch from dbt if a small team without a data engineer needs the whole flow in one place.
Check Keboola in the ranked list (1 out of 19, ETL and data integration) →
Switzerland · Published
KNIME AG is registered in Zurich and gives away an open source desktop client for building workflows by drawing them rather than coding. The hosted Team plan starts at 99 euros a month with 3 seats and 120 execution credits, and the pricing page states there is no per-seat licensing. Larger capacity is quoted.
Switch from dbt if analysts who do not write SQL should build the transformations visually.
KNIME comes 14 of 19 when ETL and data integration products are ranked here →
United Kingdom · See profile
Easy Data Transform is a Windows and Mac program from Oryx Digital Ltd in England, sold for a one-time fee from $99, 90 euros or 75 pounds plus tax, with a free trial and a 60-day money-back guarantee. Data stays on the computer unless you send it out. It handles files such as Excel, CSV, JSON and XML, not a warehouse.
Switch from dbt if the work is file wrangling on one machine and a subscription feels excessive.
More on Easy Data Transform: ranked 19 of 19 in ETL and data integration →
France / United States · Published
Dataiku, French and American, spans data preparation, visual flows, machine learning and governance. A 14-day managed cloud trial covers up to five users with everything except Govern and advanced LLM Mesh, and an installable Free Edition exists for Mac and Linux. No paid prices are published, so budgeting starts with a sales call.
Switch from dbt if transformation is one step in a wider machine learning workflow.
Dataiku review, listed at 15 of 19 ETL and data integration products →
United Kingdom · On request
Matillion pushes its jobs down into Snowflake, BigQuery or Databricks, so the warehouse does the computing, much as with dbt. The difference is the visual builder, which analysts can read without SQL. Pricing is credits consumed by task hours in annual packages with a quoted rate, and visual artefacts review less cleanly in a pull request than text does.
Switch from dbt if you want to keep warehouse-side processing but hand the design to a visual editor.
Read our review of Matillion (position 6 of 19 on the ETL and data integration list) →
United States · See profile
Power Query is Microsoft's graphical transformation engine, already inside Excel, Power BI, Power Platform and Dynamics 365. Dataflow Gen2 runs it in Microsoft Fabric with 300+ transformations and several output destinations, billed pay-as-you-go by capacity unit. It stores nothing itself, so results land in whatever product hosts it.
Switch from dbt if your reporting already lives in Power BI or Excel and you want no extra licence.
Why Microsoft Power Query and Dataflow Gen2 is #17 among 19 ETL and data integration products →
United States · Published
Azure Data Factory is priced per activity run and per integration unit hour, with the rates published, and needs no separate procurement inside an Azure agreement. Its self-hosted runtime reaches on-premise databases. Debugging a failed pipeline means reading run histories in the portal, and authoring is clunky next to a text-first tool.
Switch from dbt if you sit on Azure and also need to reach on-premise databases.
Our ETL and data integration guide puts Azure Data Factory at 10 of 19; read why →
United States · Published
AWS Glue is serverless, billed per data processing unit hour by the second with a one-minute minimum, and its catalogue connects to the rest of the AWS data stack. The visual editor generates code, yet tuning and debugging still assume Spark knowledge. Frequent small runs waste the minimum billing.
Switch from dbt if your data platform is AWS and somebody on the team knows Spark.
Full verdict on AWS Glue, which sits at number 11 of 19 for ETL and data integration →
United States · On request
Talend's data quality and profiling now sit inside Qlik's platform, sold as a quoted capacity subscription metered on volume, job runs and duration. The open source Studio line was wound down, so migrating old Studio projects is painful. It targets enterprises that must show governance evidence alongside their pipelines.
Switch from dbt if auditors want data quality evidence attached to every pipeline.
Full verdict on Qlik Talend Cloud, which sits at number 9 of 19 for ETL and data integration →
United States · Published
Informatica IDMC, now owned by Salesforce, connects to mainframes and ageing on-premise applications, with lineage and governance that satisfy auditors. A free Cloud Data Integration tier covers up to 20M rows a month, while serious work is sold in consumption units on an annual commitment that make forecasting difficult.
Switch from dbt if a large regulated estate with legacy systems has to clear procurement.
Why Informatica IDMC (Salesforce) is #12 among 19 ETL and data integration products →
Weld: It combines extraction, SQL transformation and lineage, with tiers published from $99 a month and a 14-day trial.
KNIME: The desktop client is free and open source, and the Team plan starts at 99 euros a month with 3 seats.
Easy Data Transform: It is a one-time fee from $99 for Windows and Mac, and data stays on the computer unless you send it.
Azure Data Factory: Pricing is per activity run and integration unit hour, and the self-hosted runtime reaches on-premise databases.
Microsoft Power Query and Dataflow Gen2: The same graphical engine is built into both, and Dataflow Gen2 adds Fabric destinations.
dbt remains the sensible choice when your data already lands in a warehouse and your team writes SQL. Models are plain text under version control, with tests, documentation and lineage attached, and the Developer plan lets one person start free. Its licence commitments after the merger point to continuity. If your complaint is only the seat price, check whether the Developer plan or the open source distribution under Apache 2.0 covers you before leaving.
dbt models are SQL with Jinja templating and YAML configuration, so a visual or platform tool cannot import them directly. Budget for rewriting the logic, and move the tests too, since a replacement that lacks them silently drops your data checks.
Job scheduling, CI checks on GitHub or GitLab and lineage views come from the dbt platform or your own setup. List which of these your team uses daily, then confirm the destination offers each one before you cut over.
Keep dbt running beside the new tool for one full reporting period and compare row counts on the key tables before switching dashboards.
The Developer plan is free with one seat, 3,000 successful models a month and one project, and dbt's introduction describes the open source distribution as Apache 2.0 licensed. Starter is $100 per user per month. Enterprise and Enterprise+ are quoted.
No. Its documentation describes it as a transformation tool that runs alongside ingestion tools. Weld and Keboola in our list cover loading as well.
dbt Labs announced the merger on 13 October 2025 and stated that dbt Core and Fusion keep their licences, that dbt Core stays maintained and that both products continue.
KNIME offers a free open source desktop client for visual workflows, Matillion has a visual builder, and Easy Data Transform is drag and drop on the desktop.
Weld lists tiers from $99 a month, Keboola publishes $0.14 a minute for top-ups, and KNIME lists its Team plan from 99 euros a month. Enterprise tools such as Informatica and Qlik quote prices.
Weld is Danish, Keboola Czech, KNIME Swiss and Easy Data Transform English. Dataiku is French and American. The others in the list are American or sold by American owners.
Founder of The Knowledge Engineers and of 2 Steps Ahead, an AI search visibility agency. Editorially responsible for every page on this site; research and drafting are AI-assisted.
Written, edited and fact-checked by Sebastiaan Smits. Last reviewed 5 November 2026. How we review.
