Find your next open-source project

Search by name, purpose, or keyword — or paste a GitHub repository URL

airbytehq/airbyte

Data

Open-source ELT platform that syncs data from APIs, databases and files into warehouses, lakes and AI applications.

22,329
5,441
Python
NOASSERTION
6 years old

What it does

Airbyte is an open-source data movement platform for ELT pipelines. It connects sources such as APIs, databases and files to destinations including data warehouses, data lakes and AI applications, and ships a catalog of 700+ connectors. Teams can run the platform themselves as Airbyte Core or use the managed Airbyte Cloud. The project also exposes an MCP server so AI agents such as Claude, ChatGPT, Cursor and VS Code Copilot can build pipelines, read logs and troubleshoot syncs.

It is aimed at data engineers and platform teams that need many long-tail integrations and want to customize existing connectors rather than depend entirely on a closed SaaS. Connectors can be built through the no-code Connector Builder or the low-code CDK, and syncs can be orchestrated with Airflow, Dagster, Kestra or the Airbyte API.

Quick start

  1. Choose a deployment path: either deploy Airbyte Core on your own infrastructure, or sign up for Airbyte Cloud and follow the Cloud getting started guide.
  2. Pick a source and destination from the connector catalog — for example, a database as source and a warehouse as destination.
  3. Configure the connection in the Airbyte UI, select the streams to sync and set a schedule, then run the first sync and check its logs.

To use Airbyte from an AI agent, follow the Airbyte MCP overview. Airbyte Cloud agents connect to the hosted server at https://mcp.airbyte.com/mcp and sign in with OAuth via the Cloud setup guide; self-managed deployments use the self-managed setup against the local API.

A demo app is available for exploring the interface before deploying.

Constraints and licensing

Airbyte Cloud offers features not present in the self-hosted version, including the context layer that lets AI agents read live data from supported sources and query destinations such as BigQuery and Snowflake without waiting for a sync; direct reads are read-only and Cloud-only (context layer).

The repository does not use a single license: badges show both MIT and ELv2, and licensing details live in the license file and license FAQ. A 2025 pull request titled "Migrate connectors from MIT to ELv2" was posted to Hacker News, so teams should verify connector licensing before redistribution or commercial use. Airbyte Enterprise adds security and other features on top of the open-source edition (Airbyte Enterprise). Security vulnerabilities should be reported through the vulnerability disclosure policy, not GitHub issues.

HotRepo's take

Airbyte has appeared in Hacker News discussions since 2021, when the launch thread drew 68 points and 24 comments; a 2022 "Show HN" highlighted the 300+ connector count, and the 1.0 release was noted in 2024. Those threads also capture an enduring tension: one early commenter argued the project was "not a fully open source project" because some features required payment (discussion), while an Airbyte engineer replied that connectors and core ELT functionality would remain free and open source. The later MIT-to-ELv2 connector migration discussion suggests that licensing remains a point to check rather than assume.

On HotRepo, the repository had 22,325 stars and 5,442 forks as of the 2026-11-06 Python daily snapshot, adding 9 stars that day and reaching position 7 on that daily Python list. The single snapshot supports only that day's standing, not a trend.

Written by AI from the project's README, updated . The project's own docs are the source of truth.

Hacker News

Airbyte: Simple and extensible open-source EL(T)
▲ 68 · 24 comments · January 16, 2021
Show HN: Airbyte, data integration platform with 300+ open-source connectors
▲ 29 · 13 comments · December 16, 2022
Airbyte 1.0 Released
▲ 4 · 1 comment · September 25, 2024