Airbyte Cloud

Open-source data integration platform with 350+ connectors.

Reviewed by 7wData

On this page

Publisher review

Airbyte is an open-source data integration platform designed for engineering teams seeking flexibility and cost control in ELT pipelines. Since its 2020 founding, the company has positioned itself as a transparent alternative to Fivetran: customers own the source code, choose their deployment (cloud, self-hosted, or hybrid), and pay based on data volume or reserved capacity rather than per-connector fees.

The platform ships 600+ connectors covering APIs (Salesforce, Stripe, HubSpot), databases (PostgreSQL, MongoDB), and data warehouses (Snowflake, BigQuery, Databricks). For proprietary integrations, the Connector Development Kit (CDK) lets engineers build Python-based connectors in roughly 30 minutes. Native dbt integration completes the ELT workflow by triggering transformations immediately after data lands, eliminating the need for external orchestrators.

The open-source model has trade-offs. Self-hosted deployments demand Kubernetes expertise, sustained infrastructure management, and external monitoring—typical production costs run $500–$3,000 monthly. Roughly 40% of Airbyte's connector ecosystem is community-maintained, which means variable reliability, documentation quality, and update latency. When Salesforce's API changes, Fivetran's in-house team pushes a fix within days; Airbyte depends on volunteer maintainers, which can take weeks or longer.

Cloud pricing starts at $10/month but scales with data volume: 50GB monthly easily reaches $1,250/month. The newer Pro plan offers capacity-based pricing (pay per "Data Worker"), which predicts costs for stable workloads but requires annual contracts. Open-source self-hosting is free in software cost, but infrastructure and staffing dominate the total cost of ownership.

Fivetran excels in simplicity, 24/7 connector maintenance, and enterprise support. Airbyte wins on customization, sovereignty, price predictability at scale, and developer control. In 2026, with 20,000 GitHub stars and 7,000+ production customers, Airbyte has proven product-market fit among engineering-led organizations. It remains a developer-first tool, not a business-user tool.

Get the AI & data signal, daily.

335k+ subscribers read this every morning. One email, both newsletters. Unsubscribe anytime.

How it works

  1. 600+ pre-built connectors

    Covers Salesforce, Stripe, HubSpot, PostgreSQL, MongoDB, Snowflake, BigQuery, Databricks, and long-tail sources; roughly 15% source connectors and 40% destination connectors are Airbyte-maintained; remainder are community-supported.

  2. Connector Development Kit (CDK)

    Build custom Python connectors for proprietary data sources in approximately 30 minutes without waiting for vendor maintenance cycles.

  3. Native dbt integration

    Automatically trigger dbt transformations immediately after data lands in your warehouse, completing the ELT workflow without separate orchestration.

  4. Flexible deployment

    Run on Airbyte Cloud (SaaS), self-hosted (your infrastructure), or hybrid; enables data sovereignty and compliance-driven architecture choices.

  5. Capacity-based pricing (Pro tier)

    Pay for reserved Data Workers instead of per-row volume; provides predictable costs for stable workloads but requires annual commitment.

  6. PyAirbyte

    Python SDK for programmatic pipeline control, testing, and embedding Airbyte workflows into custom orchestration tools.

Strengths and trade-offs

Strengths

  • Cost-effective at scale: self-hosted is software-free; Cloud Standard at $10/month undercuts Fivetran's per-connector model for most workloads.
  • Custom connector support via CDK: build integrations in 30 minutes, reducing dependency on vendor release cycles and enabling proprietary integrations.
  • Open-source transparency: full source code access, no vendor lock-in, and ability to contribute fixes and enhancements back to the community.

Trade-offs

  • Operational overhead for self-hosting: requires Kubernetes expertise, sustained infrastructure management, and $500–$3,000 monthly in production costs.
  • Community connector reliability: ~40% of the ecosystem is volunteer-maintained; update latency and documentation quality vary significantly across connectors.
  • Steep learning curve for non-engineers: lacks drag-and-drop simplicity; requires Python knowledge for custom connectors and SQL fluency for transformations.

Pricing context

Airbyte offers six pricing tiers. Open Source is free software (you cover $500–$3,000+ monthly infrastructure costs). Cloud Standard starts at $10/month with volume-based pricing: $15 per million API rows or $10 per GB of database data.

Cloud Plus runs $25,000+/year with capacity-based Data Workers and no volume overage charges. Pro tier (contact sales) adds role-based access control and multi-region deployment. Enterprise pricing is custom for self-hosted deployments with custom SLAs and enterprise connectors (ServiceNow, Workday, Oracle). A typical moderate workload (50M API rows + 50GB database monthly) costs roughly $1,250/month on Standard Cloud; Plus becomes cost-competitive at higher volumes if committed annually.

Getting started with Airbyte Cloud

  1. Sign up for Airbyte Cloud

    Sign up for an Airbyte Cloud account. Select your pricing tier: Standard at $10/month with pay-per-volume billing, or Plus for capacity-based pricing. Verify your email to access the workspace dashboard.

  2. Connect your data source

    Browse Airbyte's 600+ pre-built connectors and select your source: APIs like Salesforce, Stripe, or HubSpot; databases like PostgreSQL or MongoDB; or SaaS platforms. Enter your source credentials to grant Airbyte read access.

  3. Configure destination and mapping

    Select your data warehouse destination: Snowflake, BigQuery, Databricks, or another compatible platform. Choose which tables or streams to replicate. Airbyte auto-creates destination schemas; customize field mappings if needed for your schema.

  4. Run your first sync

    Trigger your first manual sync from the connection settings. Monitor the job progress and logs in real-time on the dashboard. Once complete, verify that data successfully arrived in your destination warehouse.

  5. Schedule and monitor syncs

    Configure your sync schedule: hourly, daily, weekly, or on-demand. Set up email alerts for failed or delayed syncs. Review sync history logs regularly to audit data delivery and troubleshoot any issues.

Frequently Asked Questions

What is Airbyte Cloud?

Airbyte Cloud is an open-source data integration platform for ELT pipelines, founded in 2020. It offers 600+ pre-built connectors for APIs, databases, and data warehouses. Users can deploy on Airbyte Cloud (SaaS), self-host on their infrastructure, or use hybrid setups, with pricing starting at $10 monthly.

How much does Airbyte Cloud cost?

Cloud Standard starts at $10/month with volume-based pricing: $15 per million API rows or $10 per GB of database data. Cloud Plus costs $25,000+ annually with capacity-based Data Workers and no volume overages. A typical moderate workload (50M API rows + 50GB database) runs roughly $1,250/month on Standard.

Can I build custom connectors with Airbyte?

Yes. The Connector Development Kit (CDK) enables engineers to build Python-based custom connectors for proprietary data sources in roughly 30 minutes. This eliminates waiting for vendor release cycles and reduces dependency on external vendor maintenance. Custom integrations become feasible without vendor intervention.

How does Airbyte compare to Fivetran?

Airbyte excels in customization, data sovereignty, price predictability, and developer control. Fivetran offers simplicity, 24/7 connector maintenance, and enterprise support. Airbyte's open-source model enables vendor independence and lower costs at scale; Fivetran prioritizes ease-of-use and turnkey reliability for non-technical teams.

What are the risks of self-hosting Airbyte?

Self-hosted Airbyte requires Kubernetes expertise and continuous infrastructure management, with typical production costs of $500–$3,000 monthly. Roughly 40% of connectors are community-maintained, leading to variable reliability, inconsistent documentation, and slower update cycles. When vendor APIs change, fixes depend on volunteer maintainers.

Does Airbyte integrate with dbt?

Yes. Airbyte includes native dbt integration that automatically triggers dbt transformations immediately after data lands in your data warehouse. This completes the full ELT workflow within a single integrated platform without requiring separate external orchestration tools for transformation scheduling and management.

Alternatives in this category

Integrations

Snowflake BigQuery Databricks Postgres

How Airbyte Cloud compares

Direct head-to-head against 2 competitors. Picked by 7wData.

This tool

Airbyte Cloud

Pricing
Airbyte offers six pricing tiers. Open Source is free software (you cover $500–$3,000+ monthly infrastructure costs). Cloud Standard starts at $10/month with volume-based pricing: $15 per million API rows or $10 per GB of database data. Cloud Plus runs $25,000+/year with capacity-based Data Workers and no volume overage charges. Pro tier (contact sales) adds role-based access control and multi-region deployment. Enterprise pricing is custom for self-hosted deployments with custom SLAs and enterprise connectors (ServiceNow, Workday, Oracle). A typical moderate workload (50M API rows + 50GB database monthly) costs roughly $1,250/month on Standard Cloud; Plus becomes cost-competitive at higher volumes if committed annually.
Target
Airbyte is an open-source data integration platform designed for engineering teams seeking flexibility and cost control in ELT pipelines.
Deployment
self-hosted
Strength
Cost-effective at scale: self-hosted is software-free; Cloud Standard at $10/month undercuts Fivetran's per-connector model for most workloads.
Watch for
Operational overhead for self-hosting: requires Kubernetes expertise, sustained infrastructure management, and $500–$3,000 monthly in production costs.

Fivetran

Pricing
Free tier (500K MAR); Standard from $120/month plus $2.50 per million MAR, billed per connector.
Target
Mid-market and enterprise teams wanting zero-ops, fully managed ELT with no connector maintenance.
Deployment
SaaS only, no self-hosted option.
Strength
700+ vendor-maintained connectors with automatic schema migration: no engineering intervention required when source APIs change.
Watch for
March 2025 switch to per-connector MAR billing caused 40-70% cost increases for multi-connector customers; no pooled discounts.

Matillion

Pricing
Developer tier free; Teams roughly $1,000-$2,000/month; Enterprise contact sales, verified contracts $20K-$300K+/year.
Target
Analytics engineering teams on Snowflake, Databricks, or BigQuery wanting visual pipeline building with in-warehouse transformation.
Deployment
SaaS and hybrid (enterprise tier).
Strength
Visual drag-and-drop pipeline builder that pushes all transformation compute into the warehouse, eliminating a separate transform server.
Watch for
Credit billing stacked on top of warehouse compute makes TCO hard to predict; verified reports of out-of-memory errors at scale.

User reviews

No user reviews yet. Be the first to write one.

Sources

Reporting on this tool draws on these publicly available sources.

  1. airbyte.com — Airbyte Cloud pricing tiers (Standard $10/month, Plus $25k/year, Pro capacity-based Data Workers), features and credits by tier
  2. weld.app — Comparison of deployment flexibility, cost models, connector ecosystem, customization, and support trade-offs between Fivetran and Airbyte
  3. www.integrate.io — 600+ connector breadth, operational complexity of self-hosting, community connector quality variability, and support tier limitations
  4. medium.com — Data engineer perspective on open-source strengths, CDK customization capabilities, community engagement, and relative platform maturity
  5. checkthat.ai — 2026 pricing details, real-world cost examples (Standard Cloud ~$1,250/month for moderate usage), capacity-based Pro plan mechanics