--- title: "Airbyte and Stitch alternative: Airbyte vs Stitch vs Rig" description: "Airbyte vs Stitch, compared honestly, then Rig: ingestion on open-source dlt, dbt Core and governed AI, nothing billed per row. We migrate you for free." canonical: "https://rig.so/migrate/airbyte" format: markdown --- Airbyte and Stitch alternatives # An Airbyte and Stitch alternative that doesn't stop at the load First, the straight comparison you searched for: Airbyte vs Stitch. Then where Rig fits. Ingestion on open-source dlt, dbt Core, a context layer and governed AI access, with nothing billed per row. We do the migration for free and run it beside your current pipelines until the numbers match. [Book a call](https://rig.so/book-demo?utm_source=google&utm_medium=cpc&utm_campaign=airbyte)[Airbyte vs Stitch first](#versus) 01 ## Airbyte vs Stitch: what's actually different? Both move data from your sources into a warehouse, and both stop there. The differences are in how many sources, where it runs, how it bills and who owns it. Each figure below is from the vendor's own pricing page or docs, checked 28 September 2026. | | Airbyte | Stitch | | --- | --- | --- | | Who owns it | Airbyte, independent | Qlik (via Talend, which bought Stitch in 2018; Qlik bought Talend in 2023) | | Self-host | Yes. Airbyte Core is free to run on your own infrastructure | No. Hosted only | | Licence | Connectors and platform code under Elastic License 2.0 (ELv2); protocol under MIT | Proprietary service, built on the open Singer tap and target standard | | Sources | 700+ connectors, plus a no-code Connector Builder | 130+ sources, plus an Import API for pushing your own data | | How it bills | Credits on Cloud: 6 per million API rows, 4 per GB from databases and files. Capacity-priced Pro plan | Rows replicated per month, in plan bands | | Entry price | Core $0 self-hosted. Cloud Standard from $20 a month with 5 credits, extra credits $5 each | Standard from $100 a month (5m rows) | | Database CDC | Log-based, e.g. Postgres via the write-ahead log | Log-based for MySQL, Postgres, SQL Server, Oracle, MongoDB, DynamoDB | | Transformations | Triggers dbt Cloud jobs after a sync (Cloud plans) | None. Extract and load only | | Direction | Still shipping, including Airbyte Agents and an agent MCP (May 2026) | Stitch now points new users at Qlik Talend Cloud | ### Pick Airbyte if you need a long-tail connector, want to self-host at zero licence cost, or have engineers who will build and maintain custom sources. ### Pick Stitch if you have a handful of mainstream sources, low volume and want the simplest hosted setup, and you are comfortable with where Qlik takes it next. Either way you still need the rest: something to run dbt, somewhere to define what the metrics mean, a BI tool, and a safe way to let people and AI agents query the result. That is the part Rig is for. 02 ## I searched for an open source alternative. Is Rig one? No, and it is better to hear that here than on a call. Rig is a commercial product. What is open is everything your data depends on, which is usually the reason people ask: ### Connectors: dlt Rig's ingestion is built on dlt, an open-source Python library. Every pipeline is a Python module you can read, fork and run without us. ### Models: dbt Core Transformations are dbt Core, open source. Plain SQL in a git repository you own. ### Data: your warehouse Bring your own Snowflake, BigQuery, Databricks or Redshift and the data never lives anywhere we control. For completeness: Airbyte Core is free to self-host, under the Elastic License 2.0. That is source-available rather than an OSI-approved open source licence, which matters mainly if you planned to resell it as a service. Stitch is a hosted product built on Singer, an open standard for taps and targets that Stitch sponsors. If the goal is zero licence cost and you have someone to run it, self-hosted Airbyte or plain dlt is a fair answer. If the goal is not being locked in, Rig gets you there by keeping the pipes and models open and the data in your warehouse. 03 ## How does each one bill? Not a calculator, on purpose. Airbyte's capacity plans are quote-only and Stitch prices in bands, so any single number we put against them would be partly invented. Here is what each vendor publishes, and what makes the bill move. | Plan | Billed on | Published price | What moves the bill | | --- | --- | --- | --- | | Airbyte Core | Nothing: you run it | $0 licence | Your servers, and the engineer who keeps them up | | Airbyte Cloud Standard | Credits | From $20 a month with 5 credits; $5 a credit after that. 1m API rows is 6 credits, 1 GB from a database is 4 (about $30 and $20 at list, our arithmetic) | Rows and gigabytes synced | | Airbyte Cloud Plus | Prepaid credit packages | 40 credits for $189 a month, up to 2,000 for $4,999 | Rows and gigabytes synced | | Airbyte Pro | Data Workers (capacity) | Custom quote. Airbyte says no charge based on data volume | How many syncs run at once | | Stitch | Rows per month | Standard from $100; Advanced $1,500 (100m rows); Premium $3,000 (1bn rows), the last two billed annually | Rows replicated, and which plan band that puts you in | | Rig | Platform fee plus credits for the work | Start $499 ($899 on your own warehouse); Grow $2,999; Scale $4,999 a month | Questions asked, models built, agents run. Not rows, not connectors, not seats | Airbyte and Stitch prices checked 28 September 2026 on airbyte.com/pricing, Airbyte's credits docs and stitchdata.com/pricing. The per-million-row and per-GB Airbyte figures are our arithmetic at the $5 Standard credit, not a rate Airbyte quotes. Stitch Advanced and Premium are billed annually. Rig tiers are the public ones on [the pricing page](https://rig.so/pricing), and they include the dbt runs, context layer, dashboards and AI access that Airbyte and Stitch do not. ### Rows are not a budget line On Airbyte Cloud and Stitch the bill follows how much your sources change. A backfill or a re-sync after a schema change is a line on next month's invoice. ### Free is not zero Self-hosted Airbyte has no licence fee. It has servers, upgrades, worker sizing and someone who gets paged when a sync stalls. That is usually the larger number. ### The load is the cheap half Neither tool models the data or answers questions with it. The dbt runner, BI seats and the engineer between them are the rest of the bill, and Rig covers them. ◆ ## What teams have built on Rig Three published case studies. None of them is an Airbyte or Stitch migration specifically, so we are not going to dress them up as one. They show what happens once the data is in. [Suri · D2C95%of sources centralised, in just shy of two months40+ sources connected and a solo data function doing what would usually take a 10+ person team, in their own words. Then every core metric certified against the spreadsheets the team already trusted.Read the case study](https://rig.so/case-studies/suri-metrics)[Novakid · EdTech5,000+tables in the governed context layerA BigQuery estate the commercial team could not query directly. Now sales managers and ops ask it questions in Claude through Rig MCP, with permissions enforced.Read the case study](https://rig.so/case-studies/novakid)[Birdie · HealthTech3 months → 1 dayto produce reportingFrom a small early-adopter group to most of the commercial org querying the warehouse inside Claude, in two months.Read the case study](https://rig.so/case-studies/birdie) 04 ## Do I lose my data or my history? No. Airbyte and Stitch both load into a warehouse you own, and Rig loads into the same one. Nothing is copied out, and the history they have already written stays exactly where it is. We match the existing schemas, table names and column naming, including the bookkeeping columns each tool adds, so your dbt models and dashboards keep resolving after the switch. That is the detail that decides whether a migration is invisible or a month of broken reports, so we treat it as a hard requirement. No warehouse yet, because Stitch or Airbyte was loading into something you would rather not keep? Rig can host one for you, and the pricing page shows both options. 05 ## How painful is the move, and who does the work? We do. What we need from you is roughly an hour of access setup and someone who can tell us which reports actually matter. 1. 1 ### Connection review We list every Airbyte connection or Stitch integration, what it loads and when anyone last queried it. Anything nobody reads gets flagged before we rebuild it. 2. 2 ### Parallel build We rebuild the connections on dlt, port any custom connectors or Singer taps, and bring your dbt project across, while the old pipelines keep running. 3. 3 ### Row-level parity check We compare the two outputs until they agree. You see the diff, not a promise. 4. 4 ### Cutover You switch Airbyte or Stitch off, source by source or all at once. Your call, and reversible until you do. 06 ## Will I have everything I have today? On ingestion, mostly, with two real exceptions (connector count and CDC) covered in the next section. Above ingestion, Rig does things neither tool tries to. | Capability | Airbyte | Stitch | Rig | | --- | --- | --- | --- | | Connectors | 700+ | 130+ | 300+, built on open-source dlt, and we build missing ones during the migration | | Custom sources | No-code Connector Builder, low-code YAML, Python CDK | Singer taps, Import API | dlt sources in Python, written by us or by you | | Database replication | Log-based CDC | Log-based replication | Batch and incremental. Sub-minute CDC is not there yet | | Warehouse | Loads into yours | Loads into yours | Loads into yours, or Rig hosts one for you | | Transformations | Hands off to dbt Cloud | None | dbt Core included and run for you, plus generated models over the raw tables | | Semantic layer | Not a feature | Not a feature | Context layer: table and column meaning, joins, business terms, certified metrics | | AI access | Airbyte Agents: a context store and agent MCP, a separate product | None | One governed MCP endpoint over the modelled warehouse, with row and column permissions and an audit log | | Dashboards | Not a feature | Not a feature | Included, built on the same certified metrics | | Who migrates you | You | You | We do, free, running in parallel until the numbers match | 07 ## What won't I have? Five things, stated plainly, because you would find them in week two anyway. Airbyte has more connectors: 700+ against our 300+. If you depend on something in the long tail we may need to build it. We do that during the migration at no charge, and because it is dlt you can read and run the result yourself. Airbyte Core costs nothing to license. If what you want is free software on your own servers and you have an engineer to run it, Rig is not that, and we would rather say so than pretend. Log-based CDC. Both Airbyte and Stitch replicate databases off the transaction log. Rig does batch and incremental, not sub-minute CDC. If a high-write production database needs a seconds-level recovery point, keep that one pipe where it is and move the rest. Stitch Standard starts at $100 a month. Rig starts at $499 hosted. If you only need a few low-volume sources copied into a warehouse and nothing on top, Stitch is cheaper and we will tell you that on the call. The Singer ecosystem. If you have written your own Singer taps, they do not run on Rig as they are. We port them to dlt as part of the migration, but it is a port. 08 ## What do I get that I don't have today? Airbyte and Stitch are data movers. They land raw tables and their job is finished. The gap between a landed table and an answered question is where the rest of your budget and time goes. ### Nothing billed per row or per gigabyte Airbyte Cloud charges credits per million rows and per GB. Stitch bills rows per month. A backfill, a re-sync after a schema change, a chatty API: on row pricing each of those shows up on the invoice. Rig's price does not move with rows. ### Not just the pipes Airbyte and Stitch land raw tables and stop. Rig runs dbt Core on them and builds the context layer on top: what each table and column means, how they join, which metric definition is the certified one. ### Governed AI access, built in One MCP endpoint your whole company can point Claude at, with row and column-level permissions, cost-checked SQL and a full audit log. Over modelled, documented tables, not raw ones. ### No platform to babysit The honest cost of self-hosted Airbyte is the person who upgrades it, sizes the workers and gets paged when a sync stalls. On Rig that is our job. ### We do the migration Connections, dbt models and dashboards rebuilt by us, running beside Airbyte or Stitch until the outputs agree. No charge for the build. ### Unlimited seats and connectors On every tier. Analysts, ops, finance, the CEO. Nobody is priced out of asking a question. 09 ## What if Rig shuts down? A fair question for anyone leaving an open-source tool for a commercial one. The answer is structural. You own the warehouse. The ingestion is dlt and the transformations are dbt Core, both open source, both run without us, and neither is a format only we can read. If we disappeared tomorrow, your pipelines are Python modules and your models are SQL files in a repository you control, writing into a warehouse you pay for directly. What you would lose is the context layer and the interface, not the data and not the plumbing. ## Everything else Airbyte vs Stitch: which is better? It depends what you are buying for. Airbyte has far more connectors (700+ against Stitch's 130+), a free self-hosted edition and a connector builder, and bills Cloud usage in credits per million rows or per GB. Stitch is simpler, hosted only, bills rows per month from $100, and is now part of Qlik, which is pointing new users at Qlik Talend Cloud. Neither transforms or models your data: both stop once the raw tables land. What is the best open source Airbyte alternative? If you want the open source route specifically, dlt (the Python library Rig's ingestion is built on) and Singer taps are the usual answers, and Airbyte Core itself is free to self-host under ELv2. Rig is not open source software. What is open is everything your data depends on: the connectors are dlt, the models are dbt Core, and the data sits in your own warehouse. Is Airbyte actually open source? Airbyte Core is free to self-host and the code is public. The connectors and platform code are under the Elastic License 2.0, which is source-available rather than an OSI-approved open source licence; the protocol is MIT. For most teams running it in-house that distinction does not matter. It matters if you planned to offer Airbyte as a hosted service yourself. Is Stitch being shut down? Not that Qlik has announced. Stitch still sells its plans, but its own homepage now says it is part of Qlik and encourages new users to try Qlik Talend Cloud instead. No end-of-life date is published. If you are on Stitch, the practical question is whether you want to be moved onto Qlik Talend Cloud or somewhere of your choosing. How long does a migration off Airbyte or Stitch take? A week is typical for a mid-sized stack. We rebuild your connections, dbt models and dashboards on Rig while the old pipelines keep running, compare the outputs row by row until they agree, and only then do you switch off. Larger estates with database replication run longer and are phased source by source. What happens to the data already in my warehouse? Nothing. It is your warehouse. Rig writes into the same destination, and we match the existing table and column naming so the dbt models and dashboards built on top keep resolving. History already loaded stays exactly where it is. We self-host Airbyte for free. Why would we pay for Rig? If ingestion is all you need and someone on the team is happy running it, you may not. Rig earns its price on what happens after the load: dbt run for you, a context layer that makes the tables answerable, dashboards, and governed AI access for the whole company. The comparison is Airbyte plus the engineer who runs it plus a BI tool, not Airbyte on its own. Does Rig do log-based CDC? Not at sub-minute latency, and both Airbyte and Stitch do. Batch and incremental replication are solid. If one production database needs a seconds-level recovery point, keep it on its current pipe and move everything else. What about my custom Airbyte connectors or Singer taps? We port them to dlt during the migration, free. The result is a Python module in a repository you own, which runs without Rig if you ever leave. How is Rig priced? A monthly platform fee plus credits for the work Rig does, with nothing charged per row moved and no cap on connectors or seats. The tiers and the calculator are on the pricing page. Next steps ## Start with your connection list Twenty minutes. Bring your Airbyte connections or Stitch integrations and, if you have one, last month's invoice. We'll tell you what we'd charge and what we would not bother rebuilding. If the answer is that self-hosted Airbyte or Stitch is enough for you, we'll say so. Day 0 Twenty-minute call and connection review Week 1 Rig running in parallel, nothing switched off Week 2 Row-level parity check, then your call [Book a call](https://rig.so/book-demo?utm_source=google&utm_medium=cpc&utm_campaign=airbyte) Also comparing [All comparisonsRig against the other tools people weigh it with](https://rig.so/compare)[Migrating off FivetranMAR pricing, with a calculator](https://rig.so/migrate/fivetran)[Migrating off PalantirFoundry and the Ontology, in plain terms](https://rig.so/migrate/palantir)[Semantic layer comparisonWhere the context layer sits against the rest](https://rig.so/compare/semantic-layer) Airbyte and Stitch plans, prices, connector counts and licences checked 28 September 2026 against airbyte.com, docs.airbyte.com and stitchdata.com. Vendors change these; if something here has moved, tell us and we'll fix it. Rig prices are the public tiers on [rig.so/pricing](https://rig.so/pricing). Customer figures are from Rig's published case studies. --- - [This page as HTML](https://rig.so/migrate/airbyte) - [Site map for language models](https://rig.so/llms.txt) - [API and agent documentation](https://rig.so/developers)