Use it when

You need a connector catalog broad enough to cover SaaS APIs, databases, and files without writing extraction code.
A source is missing: the no-code Connector Builder or the Python CDK makes custom connectors routine.
You want the same pipelines self-hosted for control, or managed on Airbyte Cloud, without changing tools.
Database sources should sync incrementally via log-based CDC rather than full refreshes.

Think twice when

Nobody can babysit a self-hosted deployment; operational complexity is high, and Fivetran or Hevo trade that effort for a bill.
Cost predictability matters on Cloud: volume-based pricing scales with data moved.
You need transformations in the pipeline itself; Airbyte types and dedupes only, delegating modelling to dbt.
The Elastic license matters to your legal team; it is source-available, not open source.

How it runs

Self-hosted Airbyte runs as containers, typically on Kubernetes, with syncs scheduled in the platform or by cron and triggerable over the API, Terraform, and PyAirbyte. Airbyte Cloud is the managed option, billed by volume on Standard and by capacity on Pro. Loaded data lands raw, ready for a downstream transformation framework.

Details

Compare

How Airbyte answers the questions Batch ELT turns on.

Batch ELT

How it works

Authoring model

Web UI; also REST API, Terraform provider and PyAirbyte

Incremental sync

Incremental append or append+dedupe; log-based CDC on databases

Sync frequency

Manual, scheduled or cron; hourly floor on Cloud, 15 min on Pro

Schema drift

Detect and approve, or auto-propagate fields or whole streams

Running it

Scheduling

Built in: scheduled or cron syncs, triggerable by API

Connections

Source coverage

SaaS APIs, databases and files, from a large connector catalogue

Destinations

Warehouses, lakes, databases and object storage

Custom connectors

No-code Connector Builder, or the Python CDK

Transformations

Typing and deduping only; custom dbt removed, dbt Cloud on Cloud

Cost

Billing unit

Volume on Standard; capacity via Data Workers on Pro

Alternatives in Batch ELT

View all technologies

dlt

Python library that loads messy sources into typed datasets

vs Airbyte: Open source (permissive) · Self-hosted · Managed · Operational complexity: Low · Python

dlt is an open-source Python library for the extract-and-load stage, inferring schemas, normalising nested data, and handling incremental loading — installed with pip and run wherever Python already runs.

Estuary Flow

Streaming-first data movement with batch and CDC connectors

vs Airbyte: Managed · Self-hosted · Hybrid · Operational complexity: Low · Batch · Streaming · Rust

Estuary Flow is a data movement platform built on a streaming core, capturing from databases via CDC and from SaaS APIs, then materialising continuously into warehouses and lakes rather than on a schedule.